3 ms·

This bit isn't correct. "The bank must settle the transaction with the vendor’s bank using reserves held at the central bank. If the borrower never repays the

This bit isn't correct. "The bank must settle the transaction with the vendor’s bank using reserves held at the central bank. If the borrower never repays the loan, then the bank’s reserves will not be replenished, reducing its ability to lend further." If a borrower never repays a loan, then the bank's capital is reduced (on the other side of the balance sheet from reserves) and the bank has to obtain more capital (by persuading those holding created deposits to swap them for bank bonds or shares) to maintain their risk capital ratios. Reserves have absolutely nothing to do with settling transactions. They are an optimisation of the actual process - some other entity in the banking system must take the place of the depositor in the source bank if a cross bank payment is to be made. International transactions via correspondent banking don't use reserves. They use direct bank to bank mirror transactions. It would be far better if we forgot about reserves altogether [0]. They weren't even a thing in the UK until 2005. [0]: http://www.bondeconomics.com/2022/09/bank-reserves-demystified.html http://www.bondeconomics.com/2022/09/bank-reserves-demystifi...

danuker 4y ago
> It would be far better if we forgot about reserves altogether Is this different from setting the maximum risk to capital ratio to infinity? If not, won't we have worse cycles of mania and panic? And what would stop banks from eating the world, through unfettered debt creation?
- MagnumOpus 4y ago
  It is different. Capital requirements (usually something like "a minimum of 10% of the bank's balance sheet has to be funded by equity or deferrable bonds") are far more important to a bank than reserve requirements: - if bank capital is too high, return-on-equity is lower for a given amount of profit. Hence shareholders don't want too-high capital - if bank capital is too low (from too many non-performing loans) then the bank will need recapitalisation which is expensive and bad for shareholders, and if it can't get that, it will be wiped out (either lose its licence or get bought out/nationalised/gutted) So banks are incentivised to not create debt that is too-risky -- i.e. where the ratio of potential non-performing loans is too high for the interest rate they lend at, or the interest rate of raising expensive additional capital would get too high... Reserve requirements are about liquidity (bank runs) rather than about solvency - but if bank capital can be ensured to be positive and the bank remains solvent, bank runs are much less likely in the first place, and liquidity provision and bailouts to avoid bank runs won't incur losses for the rescuing entities.
  - patrickk 4y ago
    > So banks are incentivised to not create debt that is too-risky -- i.e. where the ratio of potential non-performing loans is too high for the interest rate they lend at, or the interest rate of raising expensive additional capital would get too high... Thereoretically yes. However, 2008 moral hazard (heads I win, tails you lose) of excessive mortgage lending and securitisation put paid to that notion. Staff of bailed out banks were getting their bonuses again in 2009[1]. It also touches on the principal/agent problem, where bank executives and traders don't really care if the shareholders get wiped out and the taxpayer picks up the tab. They'll get their bonus and bounce to the next opportunity if the bank itself goes belly up. Additionally, the Justice Department refused to bring criminal charges against bank exectutives[2]. So, you'll load up on risk to make excess short term profit, maybe your bank will go bust, but you'll never face charges and may get a nice bonus next year at your bank, or bounce to the next bank. Must be nice. [1] https://abcnews.go.com/Business/story?id=8214818&page=1 https://abcnews.go.com/Business/story?id=8214818&page=1 [2] https://www.newyorker.com/news/john-cassidy/didnt-eric-holder-go-bankers https://www.newyorker.com/news/john-cassidy/didnt-eric-holde...
  - andrepd 4y ago
    > and if it can't get that, it will be wiped out (either lose its licence or get bought out/nationalised/gutted) Ahahaha, funny stuff. Actually what happens is money is pooled from non-owning taxpayers and used to recapitalise with no strings attached.
skybrian 4y ago
Yes, correspondent banking is an alternative, but it seems like in practice, the central bank payment system is pretty important in the sense that it handles more transactions?

danuker 4y ago

> It would be far better if we forgot about reserves altogether Is this different from setting the maximum risk to capital ratio to infinity? If not, won't we have worse cycles of mania and panic? And what would stop banks from eating the world, through unfettered debt creation?

MagnumOpus 4y ago
It is different. Capital requirements (usually something like "a minimum of 10% of the bank's balance sheet has to be funded by equity or deferrable bonds") are far more important to a bank than reserve requirements: - if bank capital is too high, return-on-equity is lower for a given amount of profit. Hence shareholders don't want too-high capital - if bank capital is too low (from too many non-performing loans) then the bank will need recapitalisation which is expensive and bad for shareholders, and if it can't get that, it will be wiped out (either lose its licence or get bought out/nationalised/gutted) So banks are incentivised to not create debt that is too-risky -- i.e. where the ratio of potential non-performing loans is too high for the interest rate they lend at, or the interest rate of raising expensive additional capital would get too high... Reserve requirements are about liquidity (bank runs) rather than about solvency - but if bank capital can be ensured to be positive and the bank remains solvent, bank runs are much less likely in the first place, and liquidity provision and bailouts to avoid bank runs won't incur losses for the rescuing entities.
- patrickk 4y ago
  > So banks are incentivised to not create debt that is too-risky -- i.e. where the ratio of potential non-performing loans is too high for the interest rate they lend at, or the interest rate of raising expensive additional capital would get too high... Thereoretically yes. However, 2008 moral hazard (heads I win, tails you lose) of excessive mortgage lending and securitisation put paid to that notion. Staff of bailed out banks were getting their bonuses again in 2009[1]. It also touches on the principal/agent problem, where bank executives and traders don't really care if the shareholders get wiped out and the taxpayer picks up the tab. They'll get their bonus and bounce to the next opportunity if the bank itself goes belly up. Additionally, the Justice Department refused to bring criminal charges against bank exectutives[2]. So, you'll load up on risk to make excess short term profit, maybe your bank will go bust, but you'll never face charges and may get a nice bonus next year at your bank, or bounce to the next bank. Must be nice. [1] https://abcnews.go.com/Business/story?id=8214818&page=1 https://abcnews.go.com/Business/story?id=8214818&page=1 [2] https://www.newyorker.com/news/john-cassidy/didnt-eric-holder-go-bankers https://www.newyorker.com/news/john-cassidy/didnt-eric-holde...
- andrepd 4y ago
  > and if it can't get that, it will be wiped out (either lose its licence or get bought out/nationalised/gutted) Ahahaha, funny stuff. Actually what happens is money is pooled from non-owning taxpayers and used to recapitalise with no strings attached.

skybrian 4y ago

Yes, correspondent banking is an alternative, but it seems like in practice, the central bank payment system is pretty important in the sense that it handles more transactions?