Finn Brooks: Clara, hey — rough week, but I found the thing that fixed it, which was falling down a very deep rabbit hole about how money actually gets made.
Clara Bennett: That is a specific kind of rabbit hole. How deep did you get?
Finn Brooks: Deep enough that I found a 2014 Bank of England Quarterly Bulletin — written by McLeay, Radia, and Thomas — that straight-up says commercial banks do not act as intermediaries lending out deposits. Not as a caveat, not a footnote. That is the main claim.
Clara Bennett: Hold on — and that matters because fractional reserve banking, the model most of us learned, says almost exactly the opposite. Banks collect deposits, hold a fraction as reserves, lend the rest out. That's the textbook story.
Finn Brooks: Exactly — and the Bank of England is the one calling it wrong. Which is — I mean, that's the institution! That's not a blogger!
Clara Bennett: So what we're really working out today is: where does money actually come from? And the answer turns on a distinction most people never hear — base money, which only the central bank creates, versus broad money, M1 through M3, which is primarily created by commercial banks every time they approve a loan.
Finn Brooks: And those two things are not the same size — broad money is massively larger than base money. Which, if reserves don't actually drive lending, raises this question I cannot shake: what is actually stopping banks from lending way more than they do?
Clara Bennett: That question is the whole episode, honestly. And the Bank of England paper is where we start — because once you accept that a new deposit is created by the act of lending itself, not transferred from somewhere else, the whole framework shifts.
Finn Brooks: Okay wait, 'created by the act of lending itself' — I want to slow that down because I'm not sure I actually picture what that means physically. Like, what is happening in the account?
Clara Bennett: Let me just say it plainly. A ceramics studio owner walks in, rainy Thursday afternoon, gets approved for a forty-thousand-dollar equipment loan. The bank does not go find forty thousand dollars somewhere. It types a number into a ledger. That deposit now exists in her account. It did not exist an hour ago.
Finn Brooks: It just — appears.
Clara Bennett: Two entries, simultaneously. The loan goes on as an asset — money the bank is owed. The deposit goes on as a liability — money the bank owes her. Both sides balance. And broad money just expanded by forty thousand dollars in a single accounting entry.
Finn Brooks: No, but hold on — so nobody's savings account went down? Like, my neighbor's deposit didn't shrink to fund hers?
Clara Bennett: Nobody's. That's exactly what Positive Money and the Bank of England are both pointing at — the bank is not moving someone else's savings to this borrower. Now, here's where the Deutsche Bundesbank piece lands and it's — I mean, it genuinely surprised me when I first saw it stated this directly. They confirmed that a bank's ability to make that entry is independent of whether it already holds excess reserves or deposits beforehand.
Finn Brooks: Wait — independent of its existing reserves? That breaks the whole story I had in my head.
Clara Bennett: It does. And that's the part the 2014 Bank of England bulletin is essentially underlining — McLeay, Radia, and Thomas aren't just saying banks are complicated middlemen. They're saying the 'middleman' picture is wrong at the mechanism level. The deposit isn't recycled. It's new.
Finn Brooks: So the ceramics studio's forty thousand dollars — that's genuinely new money in the economy. Not shuffled. New. Okay. That changes what I thought the Fed was even doing.
Clara Bennett: And that actually breaks something most people walk around assuming is true — the money multiplier. The textbook version: central bank creates base money, banks hold a fixed fraction as reserves, lend the rest, that fraction gets re-deposited, lent again, multiplied out mechanically. The Bank of England and the Bundesbank both describe that model as inaccurate.
Finn Brooks: Wait, both of them? Like, not just the Bank of England — the Bundesbank too said the multiplier story is wrong?
Clara Bennett: Explicitly. And here's the empirical test that I think makes it undeniable — the Federal Reserve set reserve requirements to zero in 2020. If the multiplier were the real mechanism, removing the ratio constraint entirely should have been explosive. Lending should have gone vertical.
Finn Brooks: It didn't, though. I mean — did it?
Clara Bennett: It didn't. Which tells you the reserve ratio was never actually the binding constraint.
Finn Brooks: After 2008, the Federal Reserve and the Eurosystem both ran these massive quantitative easing programs, pumping trillions into bank reserves. Trillions. And broad money, M3, barely moved proportionally. Like, if the multiplier were real, that is — that should be impossible? You added reserves at the base, the multiplier should have cascaded that upward.
Clara Bennett: That QE result is — yeah, that's the empirical coffin for the multiplier story. Banks sat on those reserves. And that points you toward endogenous money theory, which is the framework that actually fits: money supply is driven by credit demand and banks' willingness to lend, not by what the central bank injects at the base.
Finn Brooks: And there's a Philadelphia Fed researcher — Edison Yu — who found that the funding liquidity banks create through their own lending actually exceeds the liquidity that comes from deposits. Which is — wait, that means the traditional picture is literally inverted.
Clara Bennett: Inverted, yes. Deposits don't fund lending — lending generates deposits. The direction of causation runs the other way from what the multiplier assumes.
Finn Brooks: Okay but now I need to know — if reserves aren't the constraint and the multiplier is fiction, what actually stops banks from lending infinitely? Because we haven't gotten there yet and I feel like that answer is going to complicate everything we just said.
Clara Bennett: Capital requirements. That's the answer — and it completely replaces the reserve ratio story. A bank can't just keep lending because every new loan consumes capital, and regulators set a hard floor on how much capital a bank must hold against its risk-weighted assets. Run that ratio too low, you're not just in trouble with regulators — markets see it and your funding costs spike.
Finn Brooks: Wait — so it's not the Fed watching reserves, it's a capital ratio that actually bites?
Clara Bennett: In practice, yes. And the CEPR research is useful here because it pushes back on the 'thin air' framing for exactly this reason — when a bank creates that deposit, a real debt obligation comes into existence simultaneously. The borrower has to service it. If they don't, the loan goes bad, it erodes the bank's capital, and now the bank has less capacity to lend, not more. The system is self-limiting in a way the 'thin air' phrase just... doesn't capture.
Finn Brooks: No, that's — okay, that reframes it completely. It's not magic, it's a claim that has to be honored.
Clara Bennett: Right — and the other constraint that doesn't get enough credit is borrower demand. The Federal Reserve, the European Central Bank, the Bank of England — they don't directly dial up broad money. What they do is set the policy rate, which changes the cost of borrowing, which changes whether that ceramics studio owner even wants the forty-thousand-dollar loan at four percent versus eight percent. That's interest rate transmission — the real lever is over credit demand, not reserve quantities.
Finn Brooks: So the central bank is essentially — I mean, it's not printing, it's pricing. It's making the loan attractive or unattractive.
Clara Bennett: Exactly that. Base money is the central bank's domain — currency plus reserve balances. But the far larger stock of broad money? That lives or dies on whether a borrower walks through the door and whether a bank's capital position lets it say yes.
Finn Brooks: Which is why QE flooded reserves and M3 barely flinched — nobody wanted to borrow and banks were rebuilding capital. Both constraints biting at once.
Clara Bennett: Both at once, yeah. The important distinction is that unlimited money creation is theoretically available at the keystroke level but practically self-limiting — bad loans destroy the capital that enables the next loan. That's the nuance the 'thin air' framing loses, and it's the part that actually matters for understanding why the system doesn't just spiral.
Finn Brooks: The medieval goldsmiths, weirdly. Like — they figured out, what, six hundred years ago, that not everyone calls their gold at once. And that gap between what's owed and what's demanded, that's where credit gets born. That's where money gets born. And it's still — I mean, it's still the same negotiation, just now it's the IMF describing two layers, base money and broad money, and the broad layer dwarfs the base layer, and all of it runs on the same bet the goldsmiths were making.
Clara Bennett: That dual-layer framing from the IMF is — yeah, it's the thing that makes monetary policy stop feeling mysterious. Conflate base money and broad money and everything the Fed does looks like it should work differently than it does. Separate them and the whole picture clarifies. The next boom, the next recession — it really does hinge on mood more than machinery.
Finn Brooks: You came in saying you fell down a rabbit hole about how money actually gets made. I don't think I knew the answer when I said that.
Clara Bennett: And now you know it's mostly a collective agreement to keep lending. Which is either reassuring or terrifying, depending on the day.
