Banks
She Borrowed the Money to Find Out Where It Came From
Economists argued about how banks work for the better part of a century. Then somebody ran the obvious experiment, and the central bank published a bulletin agreeing with him. The textbooks have mostly not noticed.
“They approved it,” a client told me, meaning the mortgage, meaning the house. “Which is a relief, because for a while I was convinced they’d run out.”
I asked what she meant by run out.
“Well, it’s other people’s savings, isn’t it,” she said, in the tone of somebody stating the obvious to a hairdresser. “There’s only so much of it to go round.”
And I agreed with her, because that is what I had believed for forty-five years.
I have sat on the wrong side of a desk in a bank myself, in two countries, three decades apart, in uncomfortable shoes, explaining a plan to somebody who was not looking up. I remember the shape of it better than the numbers: the forms, the long pause, and then the strange anticlimax when the money simply appears in an account, as though it had been queuing somewhere.
If you had asked me where it came from, I would have said exactly what she said. From savers. The bank was holding other people’s deposits and had decided to lend me some.
That answer is wrong. Not oversimplified — wrong. And the institution that says so most plainly is the Bank of England.
3 competing theories of what a bank actually does when it lends you money
1 that survived contact with the evidence when somebody finally tested it
2014 the year the experiment was published — and the year a central bank said the same thing in its own bulletin
2 theories still being taught in a great many economics courses
Three Theories
Economists have debated how banks work for centuries. Naturally, something involving money had to become complicated.
There are three main accounts.
Financial intermediation. Banks take deposits from savers and lend that money to borrowers. Simple enough, and it is what almost everybody believes.
Fractional reserve banking. Banks lend, the money gets redeposited, is lent again, and the banking system as a whole expands the money supply through a multiplier effect. This is the version with the diagrams, and it is the one most of us met at school.
Credit creation. Banks create new money when they issue loans. They do not lend out someone else’s existing cash; they create a deposit in the borrower’s account. When the loan is repaid that money disappears again, and the interest is what the bank keeps.
Now, if banks can create money through lending, that changes how we ought to understand debt, inflation, asset prices and a good deal else besides.
So how do we settle it? We look at the evidence. Because a mathematical model can look wonderfully impressive while still being wrong, and fancy equations do not turn assumptions into facts.
This is not a philosophical dispute. It is an empirical question, answerable by examining a bank’s balance sheet while it makes a loan. If banks are intermediaries, lending should involve transferring existing funds from somewhere. If banks create credit, a new deposit should simply appear.
Which raises the obvious question: why did nobody just go and look?
Somebody Went and Looked
In 2014 the economist Richard Werner published a paper in the International Review of Financial Analysis with a title that gets straight to it: “Can banks individually create money out of nothing?”
His method was beautifully direct. He arranged with a cooperating bank — a small German cooperative, by most accounts — to borrow a substantial sum, on the condition that he could watch the bank’s internal records while it happened. Reporting puts the loan at around two hundred thousand euros, taken out by Werner himself.
Then he looked for the money.
It was not transferred from another customer’s account. It was not drawn down from the bank’s reserves. It was not borrowed from anywhere else. No existing balance anywhere was reduced.
The bank made an entry. The deposit appeared because the bank wrote it.
That is the first direct empirical test of a question economists had been arguing about since before the First World War, and the answer came back for credit creation. Werner ran a second study the following year using a different method and got the same result.
What most of us picture when we think about a loan, and what is actually involved. Photos: Unsplash
And Then the Central Bank Said It Out Loud
In the same year, the Bank of England published an article in its Quarterly Bulletin called “Money creation in the modern economy.”
It is not a pamphlet from a campaign group. It is three of the Bank’s own economists explaining, in careful institutional prose, that the majority of money in the modern economy is created by commercial banks making loans.
And then, in two sentences, it disposes of the other theories. Banks do not act simply as intermediaries, lending out deposits that savers place with them. Nor do they multiply up central bank money to create new loans and deposits.
Those are the first two theories. The ones in the textbooks. Retired, by the central bank, in its own bulletin, twelve years ago.
I find the quietness of it remarkable. There was no announcement, no reform, no correction issued to the world’s economics faculties. A bulletin appeared, said that the standard account was wrong, and the standard account carried on being taught.
“Banks do not act simply as intermediaries, lending out deposits that savers place with them — nor do they multiply up central bank money to create new loans and deposits.” — The Bank of England, 2014, retiring two centuries of teaching in a single sentence and then going back to work.
Now the Part Everyone Gets Wrong in the Other Direction
Here is where this argument usually falls apart, and I would rather hand you the corrections myself.
“Out of thin air” is not the Bank’s phrase. It belongs to the commentary that followed, and it carries an implication the Bank was careful to avoid — that the whole thing is weightless, arbitrary, a conjuring trick. It is not. A loan creates a deposit, but it also creates a corresponding debt, and the debt is extremely real to the person carrying it. Nothing has been produced from nowhere; two matching entries have been created at once.
And banks are not unconstrained. This is the most important correction and it is the one most often skipped. A bank cannot simply write itself rich. It is limited by capital requirements, by liquidity, by competition for customers, by whether the lending will actually be profitable, by whether anybody wants to borrow in the first place — and, as the Bank of England states explicitly in that same article, by the monetary policy of the central bank, which ultimately determines how much money gets created in the economy.
And Werner’s test is one bank and one loan. It is the right experiment, elegantly done, and it remains a single case that is contested in parts of the literature. It is the Bank of England’s account, arrived at independently and from the inside, that makes the conclusion hard to dismiss.
The person on the other side of the desk, who is not withdrawing anybody’s savings on your behalf. Photo: Unsplash
Why It Matters
If banks were merely moving existing savings around, then the amount of credit in an economy would be limited by how thrifty people had been. Lending would be a consequence of saving.
If banks create the money when they lend, the causation runs the other way, and a great many familiar arguments change shape. Why asset prices rise when lending is loose. Why housing booms look the way they do. Why the question of what banks choose to lend against — productive business, or existing property — turns out to matter more than the total.
And it reframes who is doing what. When a bank lends, it is not redistributing a fixed pool. It is deciding, commercially, where new money enters the economy and what it will be used to buy.
That is a considerable power to hold, and it is held almost entirely by private institutions making ordinary commercial judgments about risk and return.
Back to the Chair
I trained as an electrical engineer long before I ever ran a salon, which means I spent several years being taught to distrust a model until somebody had measured it. You can draw a beautiful circuit that cannot possibly work. The drawing is not the evidence.
What strikes me about this story is not that economists disagreed. It is that for the better part of a century, three incompatible accounts of the most basic operation in banking sat side by side in the textbooks, and the experiment that would separate them — borrow money, watch the ledger — was available to anybody with a cooperative bank manager and an afternoon.
I’ve written before about Germany’s refusal to give up cash, and about how nobody is really anti-growth. This sits underneath both. Almost every argument about money assumes a model of where it comes from, and most of us are carrying a model that the central bank has publicly disowned.
My client got her house, and her mortgage did not come out of anybody else’s savings. It came into existence when she signed, and it will go out of existence, slowly, over twenty-five years, as she pays it back. The interest is the part the bank keeps.
She is not wrong to feel that it is real. It is extremely real. It is simply not what she thought it was, and I had believed the same thing for longer than she has been alive.
So the next time somebody explains that the country cannot afford something because the money has to come from somewhere, it is worth remembering that the somewhere is, in the case of most of the money in circulation, a keyboard.
Which is not an argument for spending freely. It is an argument for knowing what the thing is before having opinions about it.
And for occasionally going and looking, which costs very little and settles rather a lot.
The Salon California Journal is a space for ideas, culture, and conversation from my chair in Brasschaat, Belgium. I write about beauty, technology, society, and the intersections between them.
