As the other commenter implied, in a fractional reserve banking system (where banks lend any amount > 0, and are not just acting as vaults), banks do indeed create money [0].

Printed dollars are necessary in an amount proportional to economic activity, and the sum of printed dollars is only loosely related to the total money supply as it affects the macroeconomy (and is becoming less relevant every year).

[0] https://en.wikipedia.org/wiki/Money_multiplier

There is no fractional reserve system or money multiplier in the modern economy though, those concepts only really apply to commodity money, not fiat money. You even say banks create money by lending, which is not how a fractional reserve system works :) A bank with zero deposits can still lend money, although they might have a problem if people want to withdraw physical currency...

https://www.bankofengland.co.uk/-/media/boe/files/quarterly-...

No, the concept still applies. Your link is attempting to describe nuance, but comes across as misleading.

The multiplier is an upper bound on money creation by banks, and it does not specify a timeline. That's a fact. Obviously, not every bank loans all the way to the minimum reserve ratio (for example, currently, the reserve ratio is zero in the US, which makes for a poor denominator). Nor does every recipient of proceeds from a loan put all of that money into a lending bank. However, the money supply is still increased via lending. The increase from the issuance of a loan is immediately reflected in total deposits.

What your link is stating is that central banks respond dynamically to the total number of bank deposits, and apply other policies (such as influencing interest rates) against that number directly, which can overpower the effect of the multiplier, if the central bank chooses. However, it does not nullify it, and it very much depends on a central bank taking that action. Your link also mentions regulations which may be specific to the UK about how banks lend, but the underlying principle of money creation remains unchanged.

From your link, wouldn't: "the multiplier being the maximum amount of commercial bank money created by a given unit of central bank money" be interpreted to mean that as soon as central bank monetizes it's debt, i.e. the money is initially created, it inherently is equal to a certain amount of commercial bank money, as long as it doesn't remain unlent? So it's not really the new creation of money, but rather the realization of it's value, upon being lent by a bank?

I'm not sure if I'm interpreting you correctly, but I think the answer is no. It's best to think of it as the creation of new money, which cannot significantly be reversed. The way the Fed prints money is by buying Treasuries, which increases the money supply in two ways:

1. It allows the government to spend more money, which ends up going to a bunch of places that can't be "unspent" (government worker salaries, contractors, equipment, etc.). That money goes into various banks and ends up getting irreversibly multiplied.

2. In order to fill bids for Treasuries, the Fed front-runs other market participants with artificially low interest rates (below market), forcing other major Treasury participants (like banks) to lower their commercial interest rates in order to do something with their deposits, resulting in more borrowing.

It should also be noted that the Fed sets the Fed Funds Rate, but my understanding is that this has a less pronounced effect. And it sets the minimum fractional reserve ratio, which as of COVID, is ZERO (infinite multiplier potential, although most big banks are pretty conservative with their reserves. Although IMO they do not need to be, since the odds of a run on the bank in an approaching-cashless society is nil).

The run-proof nature of an 'approaching-cashless society', is a facet of this I'd never considered, I suppose the point is that it'd be infinitely propped-up thanks to be so detached from any real world asset. I'll have to learn more about the other points and their relationships to money creation and supply though. I'd never thought about the reversibility of money, outside of normal means of destroying it, and I'm not completely sure if I understand it well enough to agree or disagree. That said I feel like if there's a context for these creations (or maybe realizations) that reminds me of a kind of monetary inertia and that puts me on the fence still. Always so much to learn about this kind of stuff :)

Yes, but also no.

Yes, a loan creates both a credit and a debit, and they offset exactly. In that sense, nothing is created.

But the credit spends just like cash. The debit, on the other hand, does not spend like negative cash. So in the sense of the supply of money in circulation, bank loans create money.

Wouldn’t this ‘negative cash’ be a regular financial instrument like a stock short, reverse mortgage, credit default swap, etc., which they actually can use, precisely because nothing is created by this process? It is just less liquid than hard cash, which can affect the perceived trade-able value (but that too is already factored in as expected interest).

> The total amount of money, the sum of all credits and debits, is the exact same.

Can't banks lend something like 7X more money than they have in deposits?

Only as a proportion of the total amount of money deposited in the bank. That 7x means that if $100 is deposited, the bank can loan out up to $87.5 and would have to keep $12.5, which equates to loaning out x7 the amount kept as reserve. In that scenario there still was only ever $100 dollars, but depending upon how you count the money supply the bank has now 'created' an extra $87.5 and the total supply is $187.5.

In practice it's a lot fuzzier than that, but the above is the basic concept people are usually talking about. The 7x doesn't mean the government or someone else is handing them extra dollar bills, it's just the ratio of what has to be kept in reserves vs. can be loaned out, and the amount that is loaned out effectively gets counted twice in terms of the money supply (and many more times over, since that money will get deposited into a bank and loaned out again at that 7x ratio).

If that 87.50 is deposited at the bank, then they can now loan out against that as well, no? Then, on a 100$ initial deposit, they have lended out over 100$.

This is where it’s important to get pedantic about the technical details of what precisely we mean by “money” and “lending”. If money only refers to cash in vaults or reserves at the Fed, then a commercial bank loan doesn’t create money. But it does create a claim on money, which is (more or less) perfectly convertible to money on demand and thus all economic agents treat it as money.

In theory a bank could create an infinite amount of money claims, at least up to the point where nobody wants any more at the interest rate offered. However, the bank is also subject to capital requirements which determine how much risk it can take relative to the amount of equity it has (this used to be called the “reserve ratio” but that is an outdated concept).

That 7x that is lent will be deposited by the people it is spent on. The bank can then lend 7x that and keep fabricating money as long as there are people to profitably lend too. The rate of currency growth is then correlated to velocity for how fast it’s turned over from debt to deposit to debt minus the counter flow of debts being retired.

Yes, but doesn't that strictly invalidate:

> The total amount of money, the sum of all credits and debits, is the exact same

?

As the sibling noted, the order of borrowing and deposits does not matter. The summation will always be zero.

This gets to the heart of why there can be negative dynamics in a credit cycle. If I’ve lent 1 real dollar 100 or 1000x then there is a real risk of everyone needs that dollar back today. Similarly if creditors become risk averse and stop redepositing the money into the banking system, the velocity of money will tank and the number of dollars available to lend/spend will collapse. Growing the money supply only through fractional reserve dynamics also poses risks if a systemic imbalance emerges between creditors and debtors requiring some individuals/corps to be permanent borrowers e.x the Great Depression, and Recession.

In many ways the US is using treasuries to counterbalance the above by having the federal reserve purchase treasuries which will never be paid back in real dollars. However until the pandemic the government had limited ability to place those trillions of dollars in the hands of people who needed the surplus.

The sum of all credits and debits is zero (by the double-entry accounting definition), so seems like this is tautologically true, though not very useful.

The economic principle at work is you can have $1 in the bank and loan that to 7 people sequentially (who each redeposit it), or have $7 and loan it to those same people simultaneously (who each redeposit it), but in both cases the same transactions happened ($7 out and $7 in).

The Bank of England disagrees with you, when a bank loans money it adds a figure to the creditors account (a liability) and adds an equivalent loan to its books (an asset). There's no transfer of funds from somewhere to somewhere else, that's a "common sense" concept that doesn't apply to fiat money, the same as "fractional reserve banking".

https://www.bankofengland.co.uk/-/media/boe/files/quarterly-...