How the monetary system works, part 1 – deposits, reserves and bank notes

What is happening in the economy and the markets?[1] How can financial markets – or certain important parts of them, including most notably US markets for large-market-cap equities and corporate credit – be booming when businesses are shut, millions of people are out of work and a huge wave of personal and corporate insolvencies seems inevitable? 

Most answers point to the actions of the Federal Reserve and other central banks, who have been carrying out something called quantitative easing, or ‘QE’.

 

Through QE central banks increase their balance sheets by creating the means to buy financial assets. Many describe this as printing money and worry that doing so on this massive scale must result in inflation. After all, didn’t Milton Friedman once say that “Inflation is always and everywhere a monetary phenomenon”? Haven’t we learned from Zimbabwe and Weimar Germany?

 

Others say that isn’t right: when central banks buy assets, they buy them from banks and pay for those assets with something called ‘bank reserves.’ They correctly point out that bank reserves are not money in the usual sense, because they just pile up in the banks’ accounts at the central bank and never enter the real economy.[2] Since neither the banks nor anyone else can use them to buy goods and services, how can creating bank reserves cause inflation?

 

To understand what is happening and answer these questions we need to understand the money system.

---

[1] This paper leans heavily on a number of sources on Modern Monetary Theory. These include Stephanie Kelton’s The Deficit Myth, and many episodes of The MMT Podcast with Christian Reilly and Patricia Pino, especially those involving the incomparable Steven Hail. It should be stressed that the following discussion only applies to countries that have monetary sovereignty, i.e., (roughly) issue their own currency and do not have to borrow material sums in foreign currencies. This includes countries such as the UK, US, Australia, Canada, Japan and New Zealand.

[2] See for example Jeff Snider’s ‘Bank Reserves Part 1: The Great Tease’. https://alhambrapartners.com/2018/05/08/bank-reserves-part-1-the-great-tease/

 

The best place to start the explanation is with commercial bank lending, where we come to the first incorrect picture many of us have in our heads. In the picture, banks open their doors and wait for depositors. Then, when they have deposits, they lend them out to borrowers. Banks are just intermediaries. No?

 

This is wrong. Banks do not need deposits to make loans: rather, commercial banks create deposits when they make loans.

 

When someone takes out a loan from a bank two things happen. Firstly, the bank makes a record in a spreadsheet of the amount the borrower owes, and the terms of the loan (interest rate, timing of payments, and so on). Secondly, the bank creates a new deposit: it increases the balance in the borrower’s account by an amount equal to that of the loan and makes a record of this number in another part of the same spreadsheet.  

 

A loan is an asset of the bank, because it generates cash flows for the bank: the borrower has to make payments of interest and repayments of principal. Conversely, a deposit is a liability of the bank owed to the depositor: it represents rights the depositor has against the bank. For one thing, if you have a deposit of $100, the bank has to give you $100 in bank notes when you ask for them.

 

You can exploit this claim on the bank in other ways. Let’s say you want to pay $100 to someone who has an account at the same bank. Easy: you tell the bank to debit your account and credit the recipient’s. From the bank’s perspective nothing material has changed – it still has a liability to produce $100 of bank notes. Only the beneficiary has changed.

 

But what if you want to pay someone who has an account at another bank? This is more complex.

 

Call the payer Alice and the recipient Bob. Alice’s bank is Bank A, Bob’s is Bank B.

 

Alice tells Bank A to send Bob $100. But Bank A can’t simply adjust Bob’s balance like it can Alice’s, because Bob’s account can only be adjusted by Bank B. So how does it get Bank B to make the necessary change to Bob’s account?

 

Bank B won’t just add $100 to Bob’s account because that is – remember – a liability to pay $100 in bank notes, of which Bank B has a limited supply. Bank B needs to be given an offsetting claim on someone else to pay notes to Bank B if and when Bob asks for them.

 

One way to do this is for Bank A to open an account for Bank B. Then it can debit Alice’s account and credit Bank B’s account. Now Bank B has a deposit at Bank A and so a source of bank notes it can call on to pay Bob if he demands notes.

 

But a problem arises if Bank B doesn’t trust Bank A to deliver the notes when asked. After all, in history lots of banks have proven uncreditworthy. What Bank B wants is a 100% reliable source of bank notes. Step forward the central bank: the only body that can print bank notes.

 

Each of Bank A and Bank B has an account at the central bank. The central bank credits a deposit to each account. These special deposits which banks hold at the central bank are called bank reserves. Like deposits at commercial banks they represent a right to ask for bank notes.

 

The difference is that the central bank, unlike commercial banks like Bank A and Bank B, can print bank notes in whatever amount it wants. Since reserves represent a claim to deliver bank notes on demand, and the central bank can print as many notes as it likes, it can also create as many reserves as it likes.

 

So instead of opening an account for Bank B and creating a deposit, Bank A can transfer $100 of reserves from its account at the central bank to Bank B’s account. Now Bank B has a claim on the central bank for $100 in bank notes. And this claim is 100% safe, because the central bank can print bank notes and so can never fail to deliver.

 

Note that the reserves created by the central bank never leave the central bank: they only move between the accounts of the different banks. Here we find another misleading picture: – talk of ‘reserves’ suggests banks have money ‘in reserve’ which they can ‘lend out’ to their customers. But they don’t.

 

In sum: bank reserves are a means of settling payments between banks.

 

In practice, Bank A and Bank B will be asked to make many payments to each other over any given period. Let’s assume these payments must settle once a day. At the end of each day, the amount of reserves Bank A needs to pay to Bank B is simply set off against the amount Bank B needs to pay Bank A, and each bank adjusts the balances in its own depositors’ accounts. Remember that a bank doesn’t care to whom it has liabilities, only the total amount of the liabilities compared with the size of its claim on the central bank to deliver bank notes. If the amount of reserves Bank A owes Bank B in respect of a given settlement period are the same as the amount owed to Bank A by Bank B, no transfer of reserves is needed. But if one bank needs to pay more reserves to the other than it stands to receive it will make a transfer equal to the difference.

 

This netting of payments means that in practice banks in fact do not need to have the same amount of reserves as they have deposits. Only to the extent that payments from one bank to another do not match the payments in the opposite direction is there a need to transfer reserves from one bank to the other.

 

Since the system only needs enough reserves to settle this net amount, the amount of reserves that is required in the system is very small compared with the total amount of transfers, and also very small compared with the total amount of deposits held at the banks. For this reason, the total amount of reserves in the system historically has been very small compared to the size of the balance sheets of the banks. (This has changed with QE: see below.)

 

Sometimes a bank may find it does not have enough reserves at the central bank to settle the net transactions for a given period. In this case it can ask other banks to lend it some of their surplus reserves. Banks pay interest on such loans. This creates a cost for banks and indirectly influences the rates they charge their customers.

 

Consequently, it represents a tool the central bank can use to influence bank lending: indeed, most central banks have historically regarded this rate as the key monetary policy tool. In normal times, at any moment the central bank has a target rate that it wants banks to pay on these loans, according to whether it wants the banks to do more or less lending. (In the US this is called the ‘federal funds rate’.) It attempts to hit that target rate by adjusting the amount of reserves available in the system. 

Note that for such a policy to be effective the central bank must keep the volume of reserves in the system low: if all banks have ample reserves they will not need to borrow from each other and the rate on interbank lending of reserves will fall to zero.

 

The central bank can reduce the amount of reserves in the system by having the government issue bonds to the banks and taking payment in reserves. In effect, it swaps bank reserves for government bonds. Historically, reserves did not pay interest, while government bonds did, giving the banks an incentive to swap one for the other.

 

QE has changed the functioning of this system. A central bank engages in QE by buying large volumes of financial assets (usually, but not always, government bonds) from private investors. But it is very important to note that, since a central bank can only pay for these assets using bank reserves, and only banks can have accounts with the central bank that can be credited with new reserves, central banks can only buy assets directly from banks.

 

When a central bank buys assets from a bank all that happens is a change in the levels of reserves in the system. However, the central bank can also buy assets from a non-bank by using a bank as a sort of intermediary. In this case both bank reserves and new deposits – real money in the real economy – are created. This is best explained by an example.

 

Let’s say the Bank of England wants to buy £10m of gilts from a financial institution that is not a bank – Aviva, for example. And let’s assume Aviva has an account with Barclays Bank. In order to pay for the gilts the BoE wants a new deposit of £10m to arrive in Aviva’s account at Barclays. The BoE cannot make this adjustment: only Barclays can. So the central bank directs Barclays to credit Aviva’s account with £10m. Since this creates a £10m liability on Barclays – namely, a liability to deliver either bank notes or (more likely) bank reserves to settle payments to be made in future by Aviva – Barclays will require the central bank to add £10m in reserves to Barclays’ account at the central bank before it credits Aviva’s account at Barclays.

 

In other words, in effect, Barclays buys the gilts from Aviva by exchanging them for a (newly created) deposit in Aviva’s account at Barclays, and then the central bank buys the gilts from Barclays by exchanging them for bank reserves.

Two asides at this point.

Firstly, we said above that traditionally the volume of reserves in the system was kept very low. The process of QE, as described in the previous paragraph, has transformed the picture entirely. The chart below shows the total level of reserves in the US banking system since 1959. Note that the level of reserves was almost zero until QE was launched in 2008.  

Secondly, we also said above that if central banks create too many reserves the rate of interest on interbank lending will fall to zero. This has indeed happened, meaning that central banks have deprived themselves of their key monetary policy tool. The Bank of England has therefore changed the system and now pays interest on reserve balances. This rate (known as the 'bank rate') is now the BoE's key monetary policy tool.

To revert to our main theme: it is wrong to say (as some otherwise highly educated commentators do) that when a central bank engages in QE only bank reserves are created, and therefore there can be no impact on the supply of money into the real economy. Commentators are probably misled by the fact that central banks only buy bonds directly from banks, and this results only in an increase in reserves. But sometimes the banks need the reserves precisely because they are creating bank deposits in the accounts of the non-banks that were the original owners of the bonds.

 

Now consider where the money has landed after this example of QE: in the bank account of Aviva, or strictly speaking, one of Aviva’s investment funds. What will this money be used for? The precise answer is that it depends on the rules the fund must abide by. But, since this is a fund whose job is to invest money, these rules will almost certainly limit the use of the money to investment in other financial assets. Even when the central bank buys financial assets from the public, the sellers are likely to be wealthy individuals or families who themselves will reinvest the proceeds rather than spend them.

 

The result we would expect from this type of quantitative easing is therefore an increased supply of money chasing financial assets. And note that through QE the central bank is reducing the supply of financial assets by buying them. 

Let's take the US as an example once again. In 2020 the US has issued a huge volume of Treasuries, but it has bought a lot too into its System Open Market Account (SOMA), and more still have matured and fallen out of the market. The table below shows the total figures for 2020 to the time of writing.

So on a net basis $730bn-worth of US government bonds have been taken out of the system this year. More money chasing fewer assets can have only one result: a rise in asset prices.

 

But we would not expect to see a material effect on spending, i.e., on activity in the real economy. And this, it would appear, is the explanation for a conundrum: how can it be that financial markets are booming while the economy is struggling so badly? The answer is that QE does increase the money supply in a meaningful way, but that this leads primarily to asset price inflation and does very little to encourage demand and therefore spending, which is what the real economy needs.

 

There is much more to say about the background of this system. And this only begins to hint at the implications of its functioning. But that’s probably enough for one post.