Bitcoin major fail - doesn't allow credit creation (aka deflationary currency)

[MoonShadow]

I don't think so, because: BTC Banks cannot borrow Bitcoin from a central bank to buy more time.

True, no last lender means they have a bankrun risk, so that they should treat each loan much more carefully. This is by design

Exactly. If loans can still theoretically be made and fractional reserve banking and interest can still be part of the loaning process, the fact that there is no last lender and the fact that it is a well know fact that there is a known total maximum amount of coins make banks run risks very high.

It makes them higher than what you're used to, that does not make such risk "high".  A bitcoin bank that fractionally loans out twice as much as it keeps in reserves can crediblely do so, so long as it's open and up front abou it's methods and long term CD's are it's primary method of raising capital (or some other not-on-demand deposit system).

Except that the bank can't lend out bitcoins, but only something that can be redeemed for bitcoins. Would you accept a piece of paper that says: "can be redeemed for 1 Bitcoin at the first international bank of bitcoin at any time" instead of 1 Bitcoin?

I wouldn't.

The circumstance that made fractional reserve banking possible historically was the fact that such "receipt money" was already being used for other reasons (gold too hard to store, carry around, cumbersome to transact, divide up)

With bitcoin, how would this happen? Transaction, storage cost and ease of handling of the "base metal" (bitcoin) is already very low and its divisibility high.

Does anyone feel the urge or need to bring his bitcoins to the goldsmith for safekeeping?

Again: I'm not saying it's impossible: I'm just saying I don't see any chain of developments that would make it happen.

Ah, ther is the rub.  You don't see it.  I most certainly do.  Investment banking is different than what you are used to.  That said, what if WalMart or Amazon started a bitcoin fractional bank?  Certaily some would distrust either of them, others would respect the honestly.

[kjj]

I don't think so, because: BTC Banks cannot borrow Bitcoin from a central bank to buy more time.

True, no last lender means they have a bankrun risk, so that they should treat each loan much more carefully. This is by design

Exactly. If loans can still theoretically be made and fractional reserve banking and interest can still be part of the loaning process, the fact that there is no last lender and the fact that it is a well know fact that there is a known total maximum amount of coins make banks run risks very high.

It makes them higher than what you're used to, that does not make such risk "high".  A bitcoin bank that fractionally loans out twice as much as it keeps in reserves can crediblely do so, so long as it's open and up front abou it's methods and long term CD's are it's primary method of raising capital (or some other not-on-demand deposit system).

Except that the bank can't lend out bitcoins, but only something that can be redeemed for bitcoins. Would you accept a piece of paper that says: "can be redeemed for 1 Bitcoin at the first international bank of bitcoin at any time" instead of 1 Bitcoin?

You've got it backwards.  The borrower gets actual real bitcoins.  It is the depositor that gets the IOU.

[bobitza]

I know interest can be regarded as a measure of risk and as a measure of inflation protection. With a deflationary currency, how would interest be calculated? I.e. - this project has low risk of 4% and BTC appreciates on average 5% per year ... therefore your loan interest will be -1%?!

You're guessing, and poorly.  Keep thinking and you might get there on your own.

I know the example is a poor one because you can just sit on your bitcoins and get the 5% appreciation with 0 risk involved;, but I chose it to show that the current way of thinking about interest on loans does not apply to a deflationary currency.

So basically, nobody will loan money to low risk projects?

[molecular]

Again: I'm not saying it's impossible: I'm just saying I don't see any chain of developments that would make it happen.

Ah, ther is the rub.  You don't see it.  I most certainly do.  Investment banking is different than what you are used to.  That said, what if WalMart or Amazon started a bitcoin fractional bank?  Certaily some would distrust either of them, others would respect the honestly.

Hm, I might've caught a glimpse of what you see. Can you explain in a little more detail how that Amazon fractional bank would operate to help me get a clearer picture?

[molecular]

Except that the bank can't lend out bitcoins, but only something that can be redeemed for bitcoins. Would you accept a piece of paper that says: "can be redeemed for 1 Bitcoin at the first international bank of bitcoin at any time" instead of 1 Bitcoin?

You've got it backwards.  The borrower gets actual real bitcoins.  It is the depositor that gets the IOU.

So the depositor isn't expected to use the IOU as money and it's usually not transferrable to a third part, it's just his account balance at the bank? He is still told he can withdraw all his funds at any time?

I guess I was confusing some things. What I was talking about was a bank actually making a currency backed by bitcoin and giving out more of that currency than it has bitcoin to back them with. I guess that's not fractional reserve lending at all.

The point someone else brought up, namely the risk of bank runs due to the lack of a lender of last resort is still valid, though.

[DannyHamilton]

. . .So the depositor isn't expected to use the IOU as money and it's usually not transferrable to a third part, it's just his account balance at the bank? He is still told he can withdraw all his funds at any time?. . .

Sounds like you are finally starting to understand what some people here are talking about when they mention fractional reserve.

Pretending I run a bank as an example:

Over time 3,000 customers have all deposited money with my bank.  For the sake of this example lets assume that each customer has 100 BTC on deposit.  My bank has in cold storage a total of 300,000 BTC set aside "reserved" for the purposes of honoring withdrawn funds or other debits against customer accounts.  When one of my customers makes a payment to another of my customers, I don't even have to create a blockchain transaction for the transfer.  I simply update the two accounts in my bank database.  I only have to pull from the reserve and send across the blockchain if one of my customers is transferring value to somewhere outside my bank.

Now, I come to realize that with some customers making deposits while other customers are witdrawing, my cold storage never seems to drop below 100,000 BTC.  So I loan out 10,000 BTC from my reserve to someone I have determined has a high likelihood of repaying the loan.  I still have 290,000 BTC in reserve today, and am confident that my reserve won't drop below 90,000 BTC through normal business.  I now only have a fraction of my depositors funds in reserve.  Any one customer could close out his account at any time and still be paid his full 100 BTC at any time.  For that matter, most of my customers could withdraw all their bitcoin at any time, and I'd still have enough in reserve to honor those debits.  So long as I continued to provid value to my customers and thereby gain additional customers to replace the ones who have left, there won't be a problem.

Of course if too many of my customers loose faith in my bank and all try to withdraw all their funds at once, I will obviously eventually run in to a problem where I can't honor the last 10,000 BTC of debits.  If my bank is insured by someone else who is willing to take on that risk for a small fee, then the insurer will have to pay off the last 10,000 BTC to my customers and take the loss against all the fees from all the banks they insure.  Most likely, the insurer will at that time also take ownership of the loan and become the payee, further reducing their losses as the debtor continues to pay back the 10,000 BTC plus interest.

As you can see, the person receiving the loan from the bank gets "actual" BTC.  The account holders at the bank receive (or send) "actual" BTC whenever they make transfers out of the bank.  The "IOU" is shared among all the accounts on a fractional basis that increases as money leaves the bank, and decreases as the bank takes in more funds.  Ultimately the insurer pays off the IOU from the profits they acquire from the insurance premiums collected from banks that have not failed across the entire banking industry.

Regardless, while there is still only a total of 300,000 "actual" BTC in existence withing my banking system at the time of the loan, an additional 10,000 BTC has entered circulation "out of thin air".  As all my customers still have their full balances showing in their accounts totaling 300,000, but 10,000 of it has re-entered circulation in the hands of my debtor.

The smaller the fraction of account balances that I hold in reserve, the higher the chance that customers will loose faith in my bank and I'll have to turn to my insurer to make my customers whole.

[MoonShadow]

Again: I'm not saying it's impossible: I'm just saying I don't see any chain of developments that would make it happen.

Ah, ther is the rub.  You don't see it.  I most certainly do.  Investment banking is different than what you are used to.  That said, what if WalMart or Amazon started a bitcoin fractional bank?  Certaily some would distrust either of them, others would respect the honestly.

Hm, I might've caught a glimpse of what you see. Can you explain in a little more detail how that Amazon fractional bank would operate to help me get a clearer picture?

Danny did a fine job on that front.  The question then becomes, how strong of an institution would you consider Amazon to be?  How conservative their lending practices?  The root problem with central banking isn't fractional reserve lending practices per se, but the small amount of reserve the banks are obligated to maintain, since central banking isn't a free market and pretends to insure itself.  When the banking system is ultimately supported by taxpayers, in the form of cheap insurance, they tend to push the rational limits of the 'fractional' part, favoring profit over risk.  An independent banking system, subject to the forces of the free market, might still operate in a fractional manner but not nearly to the same degree due to the risks involved.

[MoonShadow]

The point someone else brought up, namely the risk of bank runs due to the lack of a lender of last resort is still valid, though.

Yes, that is still valid, and it was exactly how the 'wild' banks in Hong Kong would keep each other honest right up until the start of WWII.  If one of the banks took things too far, the other bank owners would get wind of it, and start to amass their "notes" (if they printed any, which they did in Hong Kong, usually on recycled newspaper) and if they thought that they had enough, might start a rumor of a bank run, and actually send agents with mass numbers of notes to demand gold in person.  This is what maintained the honor among thieves in the past, and it could do so again under a free market version of bitcoin dominated fractional reserve banking.

[Richy_T]

I guess I was confusing some things. What I was talking about was a bank actually making a currency backed by bitcoin and giving out more of that currency than it has bitcoin to back them with. I guess that's not fractional reserve lending at all.

No, this is a common misconception about fractional reserve banking. R.A.Heinlein used this mistaken in one of his books and I myself believe that to be the case for a while.

FRB doesn't mean a bank can lend out more than it has on its books as deposit, only that it has to actually keep a fraction of it and can lend out the rest. Of course, the government, through the Fed (I believe) can create money to lend to the banks at crazy low interest rates so in some ways it amounts to the same. This would not be possible with Bitcoin, of course.

[kjj]

. . .So the depositor isn't expected to use the IOU as money and it's usually not transferrable to a third part, it's just his account balance at the bank? He is still told he can withdraw all his funds at any time?. . .

Sounds like you are finally starting to understand what some people here are talking about when they mention fractional reserve.

Pretending I run a bank as an example:

Over time 3,000 customers have all deposited money with my bank.  For the sake of this example lets assume that each customer has 100 BTC on deposit.  My bank has in cold storage a total of 300,000 BTC set aside "reserved" for the purposes of honoring withdrawn funds or other debits against customer accounts.  When one of my customers makes a payment to another of my customers, I don't even have to create a blockchain transaction for the transfer.  I simply update the two accounts in my bank database.  I only have to pull from the reserve and send across the blockchain if one of my customers is transferring value to somewhere outside my bank.

Now, I come to realize that with some customers making deposits while other customers are witdrawing, my cold storage never seems to drop below 100,000 BTC.  So I loan out 10,000 BTC from my reserve to someone I have determined has a high likelihood of repaying the loan.  I still have 290,000 BTC in reserve today, and am confident that my reserve won't drop below 90,000 BTC through normal business.  I now only have a fraction of my depositors funds in reserve.  Any one customer could close out his account at any time and still be paid his full 100 BTC at any time.  For that matter, most of my customers could withdraw all their bitcoin at any time, and I'd still have enough in reserve to honor those debits.  So long as I continued to provid value to my customers and thereby gain additional customers to replace the ones who have left, there won't be a problem.

Of course if too many of my customers loose faith in my bank and all try to withdraw all their funds at once, I will obviously eventually run in to a problem where I can't honor the last 10,000 BTC of debits.  If my bank is insured by someone else who is willing to take on that risk for a small fee, then the insurer will have to pay off the last 10,000 BTC to my customers and take the loss against all the fees from all the banks they insure.  Most likely, the insurer will at that time also take ownership of the loan and become the payee, further reducing their losses as the debtor continues to pay back the 10,000 BTC plus interest.

As you can see, the person receiving the loan from the bank gets "actual" BTC.  The account holders at the bank receive (or send) "actual" BTC whenever they make transfers out of the bank.  The "IOU" is shared among all the accounts on a fractional basis that increases as money leaves the bank, and decreases as the bank takes in more funds.  Ultimately the insurer pays off the IOU from the profits they acquire from the insurance premiums collected from banks that have not failed across the entire banking industry.

Regardless, while there is still only a total of 300,000 "actual" BTC in existence withing my banking system at the time of the loan, an additional 10,000 BTC has entered circulation "out of thin air".  As all my customers still have their full balances showing in their accounts totaling 300,000, but 10,000 of it has re-entered circulation in the hands of my debtor.

The smaller the fraction of account balances that I hold in reserve, the higher the chance that customers will loose faith in my bank and I'll have to turn to my insurer to make my customers whole.

This is an excellent description of banking.

If someone wants to understand modern dollar banking, this is a good starting point, just replace BTC with USD in your head as you read it, and that brings you up to around 1913.  There have been two "innovations" since then that work as a team to entirely eradicate bank runs.  The first is the Federal Reserve System which can create money at will, in any quantity whatsoever, and lend it to a bank facing a run.  The second is FDIC.  Since money is imaginary anyway, there is no reason to allow a depositor to lose any because of bank mismanagement or whatever.

[odolvlobo]

The point someone else brought up, namely the risk of bank runs due to the lack of a lender of last resort is still valid, though.

Yes, that is still valid, and it was exactly how the 'wild' banks in Hong Kong would keep each other honest right up until the start of WWII.  If one of the banks took things too far, the other bank owners would get wind of it, and start to amass their "notes" (if they printed any, which they did in Hong Kong, usually on recycled newspaper) and if they thought that they had enough, might start a rumor of a bank run, and actually send agents with mass numbers of notes to demand gold in person.  This is what maintained the honor among thieves in the past, and it could do so again under a free market version of bitcoin dominated fractional reserve banking.

I'm skeptical about this story. You are saying that banks would loan money to their competitor and then start a bank run to cause the bank to fail? That seems like a losing strategy. Not only do they lose their own money, but the customers that they would get from one less competitor would  lose their money, too.

Anyway, lender of last resort apparently doesn't do a very good job of preventing bank runs. The Federal Reserve was a lender of last resort during the depression and there were plenty of bank runs. Deposit insurance (FDIC) was created to prevent bank runs. An interesting thing to note is that deposit insurance promotes risky lending by banks, making it more important to have a lender of last resort to bail out the bank.

[myrkul]

The point someone else brought up, namely the risk of bank runs due to the lack of a lender of last resort is still valid, though.

Yes, that is still valid, and it was exactly how the 'wild' banks in Hong Kong would keep each other honest right up until the start of WWII.  If one of the banks took things too far, the other bank owners would get wind of it, and start to amass their "notes" (if they printed any, which they did in Hong Kong, usually on recycled newspaper) and if they thought that they had enough, might start a rumor of a bank run, and actually send agents with mass numbers of notes to demand gold in person.  This is what maintained the honor among thieves in the past, and it could do so again under a free market version of bitcoin dominated fractional reserve banking.

I'm skeptical about this story. You are saying that banks would loan money to their competitor and then start a bank run to cause the bank to fail? That seems like a losing strategy. Not only do they lose their own money, but the customers that they would get from one less competitor would  lose their money, too.

They keep each other honest by calling each other on their bluffs. Anybody who bluffs too hard, loses out.

A passage from Cryptonomicon (That, uhh... seems to be the entire text. Don't nobody sue me for linking to it) comes to mind....

Here's how they do it: during the normal course of business, lots of paper money will pass over the counters of (say) Chase Manhattan Bank. They'll take it into a back room and sort it, throwing into money boxes (a couple of feet square and a yard deep, with ropes on the four corners) all of the bills that were printed by (say) Bank of America in one, all of the City Bank bills into another. Then, on Friday afternoon they will bring in coolies. Each coolie, or pair of coolies, will of course have his great big long bamboo pole with him--a coolie without his pole is like a China Marine without his nickel-plated bayonet--and will poke their pole through the ropes on the corners of the box. Then one coolie will get underneath each end of the pole, hoisting the box into the air. They have to move in unison or else the box begins flailing around and everything gets out of whack. So as they head towards their destination--whatever bank whose name is printed on the bills in their box--they sing to each other, and plant their feet on the pavement in time to the music. The pole's pretty long, so they are that far apart, and they have to sing loud to hear each other, and of course each pair of coolies in the street is singing their own particular song, trying to drown out all of the others so that they don't get out of step.

So ten minutes before closing time on Friday afternoon, the doors of many banks burst open and numerous pairs of coolies march in singing, like the curtain-raiser on a fucking Broadway musical, slam their huge boxes of tattered currency down, and demand silver in exchange. All of the banks do this to each other. Sometimes, they'll all do it on the same Friday...

[MoonShadow]

The point someone else brought up, namely the risk of bank runs due to the lack of a lender of last resort is still valid, though.

Yes, that is still valid, and it was exactly how the 'wild' banks in Hong Kong would keep each other honest right up until the start of WWII.  If one of the banks took things too far, the other bank owners would get wind of it, and start to amass their "notes" (if they printed any, which they did in Hong Kong, usually on recycled newspaper) and if they thought that they had enough, might start a rumor of a bank run, and actually send agents with mass numbers of notes to demand gold in person.  This is what maintained the honor among thieves in the past, and it could do so again under a free market version of bitcoin dominated fractional reserve banking.

I'm skeptical about this story. You are saying that banks would loan money to their competitor and then start a bank run to cause the bank to fail? That seems like a losing strategy. Not only do they lose their own money, but the customers that they would get from one less competitor would  lose their money, too.

No, the banks of Hong Kong issued warehouse receipt notes for gold, and these traded on par with each other based up the claim of gold weight printed upon them.  All of the banks issued more such warehouse receipts than they had gold to cover at any one time, and all of the other bankers knew it as well.  If one was taking things too far, he would also be driving down the interest rates; thus the profits of the other banks.  so the others would accumulate the warehouse receipts of the offending bank until they had enough of them that it was unlikely that the bank would be able to honor them all at once, and start a run.  Thus running the offending banker out of business.  It was called "wildcat" banking when it was the norm in the US from 1860 or so to 1890.  It was also known as the "Free Banking Era".

Anyway, lender of last resort apparently doesn't do a very good job of preventing bank runs. The Federal Reserve was a lender of last resort during the depression and there were plenty of bank runs. Deposit insurance (FDIC) was created to prevent bank runs. An interesting thing to note is that deposit insurance promotes risky lending by banks, making it more important to have a lender of last resort to bail out the bank.

Deposit insurance is an illusion anyway, but yes, it does contribute to a circular justification.

[odolvlobo]

...  so the others would accumulate the warehouse receipts of the offending bank until they had enough of them that it was unlikely that the bank would be able to honor them all at once, and start a run.  Thus running the offending banker out of business.

This is the part I don't understand. By "accumulate", I assume you mean "pay money for", and I assume the result of starting a run is that the value of the warehouse receipts goes to 0. So, the tactic was: pay for a whole bunch of warehouse receipts and then make them worthless, right?

[molecular]

@DannyHamilton: Thanks a lot dude, for your great example.

Thanks to you guys explanations I think I have a picture of how things worked in the "free banking era" and it doesn't seem too bad to me, because banks have an incentive not do "overdo it" on their fractional reserve ratio.

What we have nowadays, though, is a bit different, right? Fiat currency isn't backed by gold and can be created at commercial banks by a magical process called credit creation (or "credit expansion"?): the bank puts a new loan into its books as an asset and at the same time increases the balance of the customers account who took the loan by the amount of the loan. New money has been created. The bank cannot do this indefinitely, though, it need some "other assets" to back up the loans. Is that correct? How exactly does this work?

[MoonShadow]

...  so the others would accumulate the warehouse receipts of the offending bank until they had enough of them that it was unlikely that the bank would be able to honor them all at once, and start a run.  Thus running the offending banker out of business.

This is the part I don't understand. By "accumulate", I assume you mean "pay money for", and I assume the result of starting a run is that the value of the warehouse receipts goes to 0. So, the tactic was: pay for a whole bunch of warehouse receipts and then make them worthless, right?

No, because the banker starting the run rumors would be first in line.  It's the depositors who heard of it late that end up screwed, and angry.  It's never the bankers you see in a mob.

[MoonShadow]

What we have nowadays, though, is a bit different, right? Fiat currency isn't backed by gold and can be created at commercial banks by a magical process called credit creation (or "credit expansion"?): the bank puts a new loan into its books as an asset and at the same time increases the balance of the customers account who took the loan by the amount of the loan. New money has been created. The bank cannot do this indefinitely, though, it need some "other assets" to back up the loans. Is that correct? How exactly does this work?

A bit over simple, but yes.

[molecular]

What we have nowadays, though, is a bit different, right? Fiat currency isn't backed by gold and can be created at commercial banks by a magical process called credit creation (or "credit expansion"?): the bank puts a new loan into its books as an asset and at the same time increases the balance of the customers account who took the loan by the amount of the loan. New money has been created. The bank cannot do this indefinitely, though, it need some "other assets" to back up the loans. Is that correct? How exactly does this work?

A bit over simple, but yes.

Simple is good, as long as it allows to draw correct conclusions ;>.

What kind of assets are required for a bank to have as reserves for the credit creation?

[bobitza]

A couple of comments on the IOUs, warehouse receipts, customer deposits and loans in a BTC economy.

First. With the current system, the IOU that you get from the bank is indistinguishable from the currency. Say you deposit 100 USD in BoA account, the bank will show that you have 100 USD, not 100 BoAD that are redeemable 1 on 1 with USD. So, that's a big difference because that will mean we need another form of currency to make BTC currency work, right? Name it banks' IOUs, vouchers, litecoins, etc.

Second. There is no such thing as customer deposits and loans with BTC because BTC is built for transactions; you cannot "deposit" in a BTC bank because it's not like you will share the private key of your account with them or they will share it with you and you can check your BTC account from time to time. You will "buy" an IOU from the bank that says you can get 1 on 1 BTCs for it (like the current system). So we go back to the first point.

I still don't understand how a loans with interest system will work in a BTC based economy. Knowing that BTC can not be created by banks via "credit expansion" and that there is a limited number of them in existence, how will the interest be paid back? With what BTCs?

[Wekkel]

I still don't understand how a loans with interest system will work in a BTC based economy. Knowing that BTC can not be created by banks via "credit expansion" and that there is a limited number of them in existence, how will the interest be paid back? With what BTCs?

Simple: produce something useful (or provide services), earn BTC and repay your loan with BTC. The closed system of BTC does not block parties from using interest.