But since we are at it: loans create deposits only in some narrow technical sense, but not in practice.

You are right than in the logical second when a bank makes a loan, they just create a pair of accounting entries out of thin air. One entry in your checking account, and one entry to record the loan.

I hope you agree that most people don't take out loan to let the money just sit in their account: they spent it.

For simplicity, let's assume the common case that the people the debtor spends their money at, don't have an account with our bank.

Our bank loses reserves, when people spend deposits. It doesn't matter whether they spend via withdrawing as cash first or whether they make a bank transfer.

A bank can not operate with negative reserves.

They get more reserves by either accepting deposits or by borrowing them.

Does this make sense to you? If not, where am I going wrong? and why do you think banks accept deposits in the first place or issue bonds in their name etc?

(For simplicity, let's assume no (positive) legal minimum reserve requirements. But also let's assume that banks need a positive level of reserves to operate; eg to pay out withdrawals and net-settle interbank transfers.

If you want to challenge those assumptions, we can do that, too. But please say so explicitly. Thanks!)

Basel III makes no mention of reserves, but it does talk about liquidity. Liquidity can be and is met through interbank lending, with the center bank filling in as necessary.

https://www.bankofengland.co.uk/quarterly-bulletin/2014/q1/m... https://en.m.wikipedia.org/wiki/Basel_III http://www.debtdeflation.com/blogs/2009/01/31/therovingcaval... https://positivemoney.org/2011/04/monetary-reform-myths-1-10...

A nice bonus link on interest rates not necessarily controlling inflation in the expected way: https://new-wayland.com/blog/interest-price-spiral/

For the same reason they pay interest on deposits: https://new-wayland.com/blog/why-banks-pay-interest-on-depos...