If you’ve just started reading this series, go to part 1 here for an introduction. This series will focus on the details of how money is created in the banking system. Part 3 will explore the process in light of the 2008 financial crisis, QE, the emergence of cryptocurrencies and the war on cash.
Fractional Reserve Banking
Modern banking operates on a principle known as fractional reserve banking. Remember how the Goldsmiths observed that people only withdrew a small fraction of their total gold holdings at any given time? Banks do the same, except not with gold. That is essentially fractional reserve banking – for each new deposit made, the bank keeps only a fraction as reserves to meet day to day demand in accordance to a ratio, known as the reserve ratio. Lets assume that the reserve ratio is 10%. For every new $100 deposit a bank receives, whether from newly created money or existing circulating money, $10 is kept aside as required reserves, to cater for daily cash withdrawals and for overnight cash settlement with the Fed and other banks in the market for reserves, which is usually electronic. The 90% left over is known as excess reserves which we learnt banks don’t like to keep around earning nothing and instead use the cash to generate interest through loaning or investing. If business is active, banks with excess reserves will make loans. If business is active but loaning is too risky, banks with excess reserves will make investments. The 90% goes on to create a new deposit somewhere else and 10% of that is kept as reserve and the process repeats until $X in initial deposit creates a total of $X/0.10 in new deposits, where 0.10 is the 10% reserve requirement in decimal form. So if a $500K mortgage was the initial deposit, it can end up creating a maximum theoretical $5m in new deposits, and a collective total of $500K is kept as reserves, distributed amongst the banking system as a whole.
What about interest rates?
At the end of every business day, all banks settle their accounts. Some banks might finish up with a surplus of funds in their reserve account while others may finish the day dry and in deficit of funds. Remember the reserve ratio? Those are reserve requirements and by law each bank must have 10% (or whatever fraction) of their total deposits in reserves (cash). So the banks ending the day in deficit will get into trouble if they don’t borrow money from banks with surplus funds to cover their gap and top up to the legal reserve requirements. Deficit banks usually borrow off surplus banks in this inter-bank market and pay what is known in the US as the Fed Funds Rate on the borrowed amount to the surplus banks. If they wish to borrow from the Fed instead from the other banks, they pay a slightly higher rate, the Discount Rate. This rate is higher to encourage banks to borrow and lend each other instead from the Fed. During a credit crunch, banks stop trading with each other in the market for reserves out of lack of trust in the quality of their balance sheets, and are willing to pay a higher premium to the Fed to park their money there. The Fed senses this and becomes lender of last resort, standing ready to lend to the banks in an unlimited capacity in order to unfreeze the inter-bank market and get money flowing between them again.
The Fed Funds rate is the crucial interest rate however, which you hear about in the news as “THE interest rate”. It is determined purely by the demand and supply of bank reserves amongst this internal inter-bank market during overnight settlements. That is why the inter-bank market, which is technically the market for bank reserves, is so important in the economy. It is the life-blood of the day-to-day banking system. Monetary policy is exacted via this market when the Fed wishes to change the quantity of reserves, which influences the interest rate. This is the channel through which the Fed keeps the interest rate steady at a particular level.
The birthplace of money: The Fed
Remember that if the Fed buys and sells to the non-bank sector, M1 or money supply is altered. If it buys and sells to the bank sector, that does not alter the money supply. So in order for monetary policy to be effective in the economy, it must change the money supply. So this means the Fed must transact with the non-bank sector.
Basically, money creation starts with the central bank and continues with commercial banks. Lets say that the Fed decides to lower the interest rate and expand the money supply. The Fed starts purchasing government securities such as bonds from primary dealers in the non-bank sector, lets say for $1m. The money it pays for these securities is not obtained from anywhere else; it is simply created out of nothing, at the stroke of a keyboard. The Fed simply credits the account of the primary dealer with $1m. As hard to believe, this is how the process begins, like a magician waving his magic wand to make a rabbit appear out of thin air. The Fed does this by double-entry accounting and drawing a liability against itself as reserves, and offsetting the liability with an asset, in the form of the purchased bond security. Newly created money now begins deposit expansion. Bank A is the dealer’s bank. Lets now examine the balance sheets after this OMO transaction:
Double entry accounting can be confusing to the layman because all transactions must balance out for each balance sheet. This conceals the chicanery of money creation. The Fed’s balance sheet looks like the exact opposite to that of Bank A; it’s liability shows Bank A’s reserve account with the Fed now sitting with a brand new $1m cheque, while its asset shows its holdings of the newly acquired Treasury (bond). Brand new money that did not exist before has now entered into bank reserves, and reserves increase by $1m. Note that money supply has not yet increased because reserves are not part of it.
Assume that the dealer does not spend all of the money immediately and holds it for a bit in the account. Bank A now keeps 10% of $1m aside as reserves into its reserve account with the Fed, as legal reserve requirements. Remember the Goldsmiths? Their trick was to print much more receipts, or claims on real gold, than actual gold itself. Modern banking works on exactly the same principle – but instead, deposits are claims on reserves. There is much more deposit money than reserve money, and if you want hard cash rather than electronic digits that represent over 97% of all money, you will need to withdraw it from an ATM, which releases vault cash to you, and vault cash is part of reserves. So deposits are a claim on reserves in a similar fashion that receipts are a claim on gold. This is a very important concept.
Money creation continues with…commercial banks
As Bank A keeps $100K as reserves, or 10% of the newly created $1m, here is what happens next. For money supply to increase and multiple deposit expansion to continue, banks have to lend or invest; creating brand new deposit money that must match its excess reserves. This is the most important point to understand, because excess reserves must be replicated with deposits for money supply to increase, since reserves are not counted as part of M2 money supply. The $900K is excess reserves. Bank A creates an equal deposit of $900K out of thin air, to match its $900K in excess reserves,. It could do this by issuing a $900K mortgage loan to you, or purchasing bonds or stocks with $900K. Nothing happens in terms of deposit expansion until you, the borrower, spends the money. After making the loan, Bank A’s balance sheet will look like this:
This $900K is newly created money out of nothing by Bank A, on top of the already created $1m by the Fed, so money supply has now increased by $1,9m. Lets say you pay the vendor the full $900K to his bank, Bank B, the following day. Bank A will tentatively credit the vendor with a deposit of $900K, while waiting for the transaction to clear. The clearing bank is usually the Fed itself. At the end of each working day, if Bank A has sufficient funds in its reserve account with the Fed, the money will clear and $900K will be transferred from the reserve account with the Fed of Bank A to the reserve account with the Fed of Bank B. A bank’s own reserves with the Fed exist as funds on demand that it must tap to cover its own purchases and loan making. Bank B will then remove the hold on the new deposit to the vendor and complete the transaction. If Bank A had insufficient reserves to clear the transaction, it will borrow reserves in the inter-bank market from another bank with a surplus of reserves. So every time you get a loan from the bank, it increases money supply, provided the funds keep flowing throughout the banking system in this manner.
Wait what, – so how does money supply increase here? The original $900K in excess reserves sitting in the reserve account of Bank A just had a new “claim” made on it by your new mortgage deposit. Recall how deposits are claims on reserves as receipts are claims on real gold, and there are always more claims on real assets than real assets themselves. Before you wired the money to the vendor, a brand new deposit is created in the money supply, – your $900K mortgage deposit that you will pay the vendor. As you wire the funds, the corresponding $900K in excess reserves is transferred from Bank A’s reserve account with the Fed to Bank B’s reserve account with the Fed. It is separate from the $900K deposit that Bank B will inherit and Bank A will lose. The important thing to realise is that the banking system overall has now $900K more than it did before the mortgage loan. All deposit transfers have a corresponding reserve transfer that’s go with it. Here is what the balance sheets of Bank A and Bank B look like after this transaction:
Shortly after the Fed has cleared the transaction, Bank B just incurred a $900K deposit liability for the vendor (all deposits are liabilities for banks because they are liable at any time to provide funds to their customers to cover their spending and withdrawals) and an equal amount of new reserves. Notice that Bank A loses the deposit money and an equal amount in reserves as they go to Bank B. Bank B upon receiving the new funds needs to hold only 10% ($90K) as required reserves in the reserve account with the Fed to cover the new deposit liability. Bank A no longer has the liability of $900K but it also has lost that much in reserves.
After Bank B deducts $90K in required reserves and leaves it in its reserve account with the Fed, it ends up with $810K in excess reserves. To match excess reserves with deposits, it decides to make a purchase of bond assets, effectively crediting the seller’s account with fresh new money to the tune of $810K, at Bank C. I don’t know why I labelled the seller a stockbroker when he is selling bonds. The balance sheets of both banks now look like this:
Bank C will then take 10% of its new $810K reserve amount as required reserves and have $729K left to lend or invest. If Bank C decides to say, lend to a small business, it will create a brand new deposit of $729K in the bank account of the business owner, matching its excess reserves. It then transfers exactly $729K in reserves to the business owner’s bank, Bank D, and the process continues. If you want to calculate the maximum amount of bank deposit money (i.e claims on reserve cash) that can be created by this process, just sum up all the bank assets in the process.
The multiple deposit expansion process continues until the original $1m Fed money theoretically creates a maximum additional $9m in commercial bank deposit money. This is the money multiplier. An initial open market operation of $1m, in ‘high powered money’ initiated by the Fed, ends up adding an extra $9m in deposits created by commercial banks through lending and investing in accordance to excess reserves. Carrying the process to its limits, excess reserves in the system drop to $0, that is, they are exhausted through lending and investing. Required reserves always equal the initial deposit of $1m, and the loaning and investing create $9m in deposits. In the long run, depending on business and economic conditions, the aggregate banking system could see the following maximum theoretical effects:
Note the word ‘maximum’ and ‘theoretical’. Much less than $9m could end up being created due to imperfections and blockages along the way such as for example, a loss in confidence, partial (and not full) spending of deposits, tighter lending conditions, financial crises and bank runs. We described the ideal conditions where the maximum amount of money could theoretically be created if the process continues smoothly.
And that is the mechanics of money creation in modern banking.
Modern Money Mechanics (A Federal Reserve Publication)
I Bet You Thought (A Federal Reserve Publication)