Where Money Comes From

Most people picture money as a pile of notes somewhere and banks as middlemen who hand it out. Both pictures are wrong. Money is a liability that appears when a bank writes a loan — and knowing which layer you are looking at changes every policy argument you will read.

Money sits on a balance sheet

Ask where money comes from and most people describe a pile: notes in a vault, deposits parked in a bank, waiting to be lent onward. There is no such pile.

Every unit of money is a liability of whoever issued it and an asset of whoever holds it. A banknote is a liability of the central bank. The balance sitting in your account is a liability of your bank toward you. When you look at your money, you are looking at one half of a ledger.

That is the whole trick of the money question. Money is not a substance moving around a container. It is a claim, and a claim exists only as two matching entries — the issuer's debt and the holder's asset. The entry is the money.

Loans create deposits, not the reverse

The textbook story runs deposits first and loans second: savers put money in, the bank lends most of it out again. The causal order is backwards.

Read a new loan as a two-column ledger. Left column, assets: the loan — the borrower's promise to repay, which the bank now owns. Right column, liabilities: the deposit — the spendable balance the bank now owes the borrower. Both entries are written at the same instant, by the same act.

Nothing was moved out of anyone's savings. The loan did not come from a deposit; the deposit exists because the loan does. What the bank needed was a borrower it expected to repay and enough capital to absorb the loss if that judgement was wrong — not a stockpile of pre-existing deposits.

This is why the term credit creation is used rather than "lending out money". The bank creates the deposit it lends. Making the loan and making the money are one operation, not two.

Two layers, not one

Once money is a liability, the next question is: a liability of whom? There are two different answers, and running them together wrecks most monetary arguments.

Base money is the liability of the central bank — the notes in circulation plus bank reserves, the deposits that commercial banks themselves hold at the central bank. You cannot spend reserves. Only banks hold them, and they hold them to settle payments with each other and with the government.

Broad money is the deposit money the public actually holds and spends — the figure in your account, and the one that appears in the money statistics. It is a liability of commercial banks, not of the central bank.

So these are not one quantity drawn at two sizes. They are two different claims on two different issuers, and they do not move together. A central bank can expand its own liabilities a great deal while the money the public spends barely moves — and the reverse can happen too.

What a central bank actually controls

A central bank does not hand out a fixed quantity of reserves and let the price of them float. It does the opposite: it sets the price and supplies the quantity.

That price is the policy rate, and it is set on a very short-dated market — banks borrowing and lending reserves among themselves overnight, to square up the day's payments. The central bank holds that rate at its target by supplying whatever quantity of reserves the system needs at that price.

Notice what that sentence claims. Reserve supply is elastic. If the banking system needs more reserves to clear its payments, it gets them; the interest rate is what is pinned down, not the quantity of reserves.

And it does not multiply a fixed stock of reserves into a fixed stock of lending. Nothing on the central bank's balance sheet forces a bank to lend, and nothing on it stops a bank from lending either. Policy works through the price of credit and the demand for it, not through a rationed input that must be spread out.

Quantitative easing is a swap

Quantitative easing is a central bank buying bonds and paying for them with reserves it creates. The seller of the bond ends up holding a deposit at the central bank instead of a bond.

Read the two sides. The private seller gave up a bond and received a reserve claim. The central bank took the bond and issued a liability. Net private wealth did not change. What changed is composition — and the duration of what the private sector is holding.

That is what gets lost when the operation is described as printing money or as a gift. It is an exchange of one safe claim for another at a price the central bank chooses. It can push bond prices up and yields down, and it moves interest-rate risk from private balance sheets onto the central bank's, without creating a single unit of net wealth.

The layers picture makes the boundary visible. This kind of operation lives almost entirely in the inner layer: reserves grow against the central bank while the outer layer — the money the public spends — moves only if banks and their borrowers do something with the change.

Why the money multiplier is taught backwards

The money multiplier is presented as a machine: feed in a quantity of reserves, apply a ratio, and out comes a quantity of deposits. Reserves first, lending second.

That inverts the causality. A bank lends when it finds a borrower worth lending to. When the resulting payments have to settle, the banking system needs reserves — and the central bank, which has committed to a rate target, supplies them.

So reserves are supplied on demand, after the lending, not before it. The multiplier states a ratio that holds once everything has settled. Taught as the engine of lending, it is a description read backwards and called a cause.

Keep the test to hand for any money claim you meet: which layer is it about, whose liability is it, and was the quantity fixed or the price? Answer those three and the claim usually resolves itself.

Related: Bank Reserves · Credit Creation · Quantitative Easing · Money Multiplier

Next: The Price of Time