If you have 100 bitcoins in your computer wallet and 100 in your MtGox account,

[BurtW]

Great!  So we all agree:  debt instruments denominated in BTC are not, in fact, BTC.  To the extent these instruments are used as money, lending increases the supply of BTC denominated money.

[1713584588]

1713584588

[1713584588]

1713584588

[1713584588]

1713584588

[1713584588]

1713584588

[DeathAndTaxes]

I'm the author of the "pure nonsense" edits. While I admit that my original posts were not very intelligible, I maintain that my position is correct.

Merely by depositing Bitcoins into Mt. Gox, and Mt. Gox lending out a proportion of them to someone else, the money supply is not affected. It only would be affected it someone accepted the Bitcoin deposit as a method of payment. At the moment, this is only even possible though the redeemable "Mt.Gox code" and the acceptance thereof is limited.

That is not true.  How do you think fractional reserve banking works?

Lets look at USD first (because everyone is familiar with it).
You deposit $100 in Bank of America.  
They lend $50 out.  The money supply has increased $50.  
You need to pay a bill and withdraw $10.  
BofA gives you $10.

BofA doesn't issue an alternate script, a BofA token.  They give you USD which are just as accepted as the USD you deposited and the USD they lend out.

You keep going back to this belief that Mt. Gox (or any "bitcoin bank") would need to issue some alternate currency/script/note and then that note be accepted as equivelent to BTC.  Where do you get that idea from?   I believe that misconception is the basis for all of your misinformation.

Lets modify the BofA example for Bitcoin.... BitcoinInternational Bank.

You deposit 100 BTC @ BIB.
BIB loans 50 BTC to a borrower.
Money supply has expanded 50 BTC.
You need to pay a bill and withdraw 10 BTC.
BIB gives you 10 BTC (not some alternate bank script, they give you 10 BTC in a normal Bitcoin transaction).

Simplisticly this is the entire basis of fractional reserves.

The difference between FRB with gold/fiat and Bitcoin is that the derivative media of exchange (hypothetical Bitcoin-substitutes) do not provide general advantages over Bitcoin itself, and therefore there is little demand for them. People might use bank notes because gold coins are too heavy, or EFT because they are not at the same location. This creates a demand for exchanging these claims instead of the original gold (for example). With Bitcoin, these problems do not exist, because the Bitcoin value of a storage medium does not affect the weight of it, and you can transfer Bitcoin electronically without an intermediary. So there is little demand for the features which normally require banks.

That is a whole different argument.  First of all lets drop this bitcoin-substitutes.  Does BofA issue USD substitutes or do they lend out USD, accept USD deposits, and return USD deposits.  A bitcoin bank would do exactly the same thing.

The question of demand for Bitcoin banks is a genuine question.  There may not be much need for Bitcoin banks thus the amount of deposits would be low and the MONEY MULTIPLIER for Bitcoin economy would also remain low.  However that is different from saying it is impossible to have fractional reserves.  

Demand for going to the moon is low but it isn't impossible to go to the moon.  

We don't know how much end user demand will eventually be created however that doesn't make fractional reserves impossible.  Convenience is only one reason people use banks.  Security is another.  The ability to earn yield is another.  

[Steve]

OK, so Steve and others do not like everyone else's definition of the phrase "money supply" so they make up their own.  There are already several definitions of money and money supply so I guess a few more will not hurt anything.  However, I have done my best and I personally am done arguing over the definition of a phrase that was ambiguous to begin with.

I think the use of "monetary base" to refer to the 21million bitcoins is useful.  Note, I didn't make up any definition of money supply.  The term itself is ambiguous as evidenced by the fact that every central bank publishes many different measures of "money supply" (M0, M1, M2, etc).

It's probably more useful to think in terms of "monetary base" and lending activity rather than try and come up with different measures for a money supply.  You have 21 million BTC of monetary base, that's it.  And then you have various forms of lending, some of which might serve as near substitutes for actual bitcoins, but which are still definitely not bitcoins.  What matters is the level of lending activity and the liquidity of those loan instruments (the higher the liquidity of a particular loan agreement, the more it is able to function like money).  

In the bitcoin world, we'll never likely see anything be a perfect substitute for actual bitcoins due to inelastic supply (just like nothing is a perfect substitute for physical gold).  In the US, since you have the FDIC backing most demand deposits, people can safely treat them as perfect substitutes for physical money (safe in the sense that they'll be worth what any other dollar is worth, but not so safe in terms of the value of those dollars).  If an exchange existed that traded mtgox BTC deposits, they would trade at a small discount to actual bitcoins to account for the counter-party risk associated with those deposits (unlike FDIC backed deposits of USD which would be just as valuable as any physical dollars).

[DeathAndTaxes]

It's probably more useful to think in terms of "monetary base" and lending activity rather than try and come up with different measures for a money supply.  You have 21 million BTC of monetary base, that's it.  And then you have various forms of lending, some of which might serve as near substitutes for actual bitcoins, but which are still definitely not bitcoins.  What matters is the level of lending activity and the liquidity of those loan instruments (the higher the liquidity of a particular loan agreement, the more it is able to function like money). 

Lending Bitcoins is Bitcoins just like lending USD is dollars.

You deposit $100 at a US bank.
The bank lends $50.
The money supply has increased $50.
You need to pay a bill and withdraw $10.  It isn't the same $10 you deposited but it is just as accepted.
The bank didn't issue some alternate currency for withdraws it gave you a "genuine" USD.

You deposit 100 BTC at a Bitcoin bank.
The bank lends 50 BTC
The money supply has increased 50 BTC.
You need to pay a bill and withdraw 10 BTC.  It isn't the same 10 BTC you deposited but it is just as accepted.
The bank didn't issue some alternate currency for withdraws, it gave you a "genuine" 10 BTC.

There is no need to "trade" demand deposits.  Have you seen anyone trading Bank Of America savings accounts, or Chase checking accounts.  "I got a checking account worth $280 Chase dollars (not real USD because they are in a Chase bank),  I will accept $0.90 cash for $1.00 deposited.".

Of course not.  Nobody trades demand deposits.  Anywhere on the planet.  How MUCH money is deposited and thus available for lending determines the money multiplier.  If trust of Bitcoin banks is low then only a small amount will be deposited and thus money multiplier will be low.  However it still does increase the money supply when used.

[Steve]

BofA doesn't issue an alternate script, a BofA token.

Yes they did…it's called a BofA bank account balance.  If that balance doesn't happen to be covered by FDIC, would you treat it as a perfect substitute for physical dollars?  No, you wouldn't (or at least you shouldn't) especially considering the ill-health that BofA is in right now.  Substitute MF Global for BofA and the point becomes much clearer.  Now, if the account is FDIC backed, it's a completely different story because you know your balance is protected in the event of a bank failure (because the FDIC has access to the printing press if it needs it).  In that case the bank account balance is a near perfect substitute for base money.

[lonelyminer (Peter Šurda)]

Merely by depositing Bitcoins into Mt. Gox, and Mt. Gox lending out a proportion of them to someone else, the money supply is not affected. It only would be affected it someone accepted the Bitcoin deposit as a method of payment. At the moment, this is only even possible though the redeemable "Mt.Gox code" and the acceptance thereof is limited.

That is 100% not true.

Kindly review the quotes I referenced. Here they are again:

Wikipedia on demand deposits :

Demand deposits are usually considered part of the money supply, as they can be used, via checks and drafts, as a means of payment for goods and services and to settle debts.

Another link: http://www.economicsjunkie.com/true-money-supply/ :

Virtually everyone accepts payment in demand deposit money. Demand deposits are thus to be included in the money supply.

Rothbard in Austrian Definitions of the Supply of Money:

It is important to recognize that demand deposits are not automatically part of the money supply by virtue of their very existence; they continue as equivalent to money only so long as the subjective estimates of the sellers of goods on the market think that they are so equivalent and accept them as such in exchange.

Where are your references?

[finway]

In the Fiat banking system, usually the money multiplier from m0 to m1 is 8 times.
But MtGox's BTC money multiplier is infinite, as long as nobody withdraw BTCs from it.

[Steve]

You deposit 100 BTC at a Bitcoin bank.
The bank lends 50 BTC
The money supply has increased 50 BTC.  <--- only to the extent that the 100 BTC deposit is considered a near perfect substitute for bitcoins
You need to pay a bill and withdraw 10 60 BTC.  It isn't the same 10 60 BTC you deposited but it is just as accepted.  Really?
The bank didn't issue some alternate currency for withdraws, it gave you a "genuine" 10 50 BTC and declared bankruptcy.

Slightly modified scenario.

[DeathAndTaxes]

In the Fiat banking system, usually the money multiplier from m0 to m1 is 8 times.
But MtGox's BTC money multiplier is infinite, as long as nobody withdraw BTCs from it.

For a short time but eventually a run of the bank could collapse it and the money multiplier would shrink to 1x.  In a free market the effective money multiplier for the economy will be set by individuals.

Overall Money Multiplier = (% of monetary base subject to fractional reserves) * (1/fractional reserve used by banks on average) + (% of monetary base outside of fractional reserves) * (1).

Thus if say hypothetically only 10% of bitcoins end up in Bitcoin banks and those banks on average have a 50% fractional reserve then the money multiplier would be a very small (0.1 * 1/0.5) +0.9*1 = 1.1 

[lonelyminer (Peter Šurda)]

The money supply has increased 50 BTC.  <--- only to the extent that the 100 BTC deposit is considered a near perfect substitute for bitcoins

Good to see that someone seems to understand what I'm talking about  .

[lonelyminer (Peter Šurda)]

In the US, since you have the FDIC backing most demand deposits, people can safely treat them as perfect substitutes for physical money (safe in the sense that they'll be worth what any other dollar is worth, but not so safe in terms of the value of those dollars).

This is a very interesting point. I'll borrow it if you don't mind

[BurtW]

Hello,

I'm the author of the "pure nonsense" edits. While I admit that my original posts were not very intelligible, I maintain that my position is correct.

Kindly review the bottom of the wiki talk page that quotes examples by mainstream as well as Austrian authors, that claim:

• Money supply is the sum nominal value of instruments that are used as media of exchange, not the sum of the nominal value of zero-maturity instruments.
• The reason why demand deposits usually appear in the money supply is that they are used in exchange instead of the reserves. If they are not used in exchange, they are not a part of the money supply.

The above bolded statement is the crux of this thread.  If you agree with it you are on one side, if you disagree with it you are on the other.  The first side, those that agree, are defining money as having only one use case - as a medium of exchange.  The other side, those that disagree with the statement are defining money as having two use cases:  1) as a medium of exchange and 2) as a way to store value.

If you believe that money is a way to store value then you believe your deposits are money.  Those that disagree are saying that deposits are not money or more specifically static deposits are not counted as part of the money supply.

[DeathAndTaxes]

You deposit 100 BTC at a Bitcoin bank.
The bank lends 50 BTC
The money supply has increased 50 BTC.  <--- only to the extent that the 100 BTC deposit is considered a near perfect substitute for bitcoins
You need to pay a bill and withdraw 10 60 BTC.  It isn't the same 10 60 BTC you deposited but it is just as accepted.  Really?
The bank didn't issue some alternate currency for withdraws, it gave you a "genuine" 10 50 BTC and declared bankruptcy.

Slightly modified scenario.

Your first modifications is dubious.  If someone someone doesn't think BTC on deposit has the same value then ... they won't deposit.  If they do deposit then that is a manifestation of the belief that the deposit is a substitute for "actual BTC".   So while a Bitcoin bank may find it harder to attract deposits any deposits it does attract are part of the money supply.  Nobody is arguing that Bitcoin likely won't have a LOWER money multiplier but that is a far different thing than saying the money supply = 21M and can't change (money multiplier of 1x into perpetuity).

The second modification is merely a run on the bank scenario. The same change to the USD bank scenario would have had the same outcome.   The scenario was intentionally simplistic.  In reality the bank would have some or all of this elements:

a) have more depositors so the action of any individual depositor would be insignificant
b) have access to short term funding from other entities (bank being a creditor) albeit without a central bank to be lender of last resort
c) have ability to sell loans as reserves fall to ensure it has sufficient assets available for demand
d) sell additional stock to raise reserves.
e) issue bonds in the name of the bank to raise reserves.

Exactly like a USD bank.

So:
1) BTC can be used in fractional reserve
2) The money supply thus can be >21M
3) The monetary base can never exceed 21M so that puts an "effective limit" on the money supply at 21M * (Effective Money Multiplier).
4) Due to inherent differences between fiat money and Bitcoins the money multiplier for BTC economy will likely be lower.

[lonelyminer (Peter Šurda)]

Money supply is the sum nominal value of instruments that are used as media of exchange, not the sum of the nominal value of zero-maturity instruments.

The above bolded statement is the crux of this thread.  If you agree with it you are on one side, if you disagree with it you are on the other.  The first side, those that agree, are defining money as having only one use case - as a medium of exchange.  The other side, those that disagree with the statement are defining money as having two use cases:  1) as a medium of exchange and 2) as a way to store value.

Please review the references:

Definition of money supply from Merriam-Webster:

the total amount of money available in an economy for spending...

Another definition:

The total supply of money in circulation in a given country's economy at a given time.

Austrian definition of true money supply:

Algebraically, TMS = Standard Money (held by the public) + Money Substitutes

The money supply does not measure store of value of everything that is listed as denominated in the money unit, but what's available for exchange.

[DeathAndTaxes]

Demand deposits ARE available for exchange.
I have $100 in a USD bank.  I realize I need to pay a bill. I withdraw $10.  The $100 was available for exchange.

Demand USD deposits are considered part of the USD money supply.
Demand BTC deposits would be considered part of the BTC money supply.
There is no difference between BTC & USD with that respect.  None.

[lonelyminer (Peter Šurda)]

Your first modifications is dubious.  If someone someone doesn't think BTC on deposit has the same value then ... they won't deposit.

Value is subjective. Merely because a depositor thinks that the deposit has value, it does not follow that other people think that too. If they are not a customer of the same issuer, they might just as well see this derivative instrument as an additional cost.

[Steve]

I mostly agree (and I think most people here are in broad agreement)...I really enjoy this thread.

Your first modifications is dubious.  If someone someone doesn't think BTC on deposit has the same value then ... they won't deposit.  

Sure they would…I don't consider a mtgox deposit to have the same value as BTC over any substantial length of time. But in the very short run I consider it more valuable than BTC because it can enable me to exchange those BTC for dollar deposits.  The value of a mtgox BTC deposit can fluctuate independently of the actual BTC it represents.

The second modification is merely a run on the bank scenario. The same change to the USD bank scenario would have had the same outcome.

Not exactly…if it were an FDIC insured account, the depositor would not lose anything.  In the BTC case, the depositor would lose actual bitcoins because there is no FDIC backing up the account.  The FDIC backing increases the value of the deposits to the point were is makes those deposits a near perfect substitute.

2) The money supply thus can be >21M

Yes, but I quibble with the notion that there is any single definition of money supply.  We're treating lending instruments (of which demand deposits are but one example) as if they are either a substitute or not a substitute (and using that as a test of whether the are included in the "money supply") but the reality is that they can be nearly a substitute, almost a substitute, somewhat a substitute, a substitute under a broad definition of money supply, or not a substitute at all.  And all this really says is that lending instruments have a value and liquidity that can fluctuate independently of the underlying BTC they represent…some are almost as good as the underlying BTC, and some are not.  Those things there are nearly as valuable and liquid (spendable) as BTC would be counted as part of narrower definitions of money supply…things that are less valuable and/or liquid would only be counted under broader definitions of money supply or not counted at all.

4) Due to inherent differences between fiat money and Bitcoins the money multiplier for BTC economy will likely be lower.

I don't know if I would make that assumption (it may or may not end up that way)…I can imagine scenarios involving a ripple-like debt network where p2p, short term lending would be far more common…you would have more adhoc, informal lending on a p2p basis where trust is more important than formal risk assessment…and you would also have your more traditional lending where risk assessment is rigorous.  Under such circumstances, it's conceivable that the multiplier could actually be greater.

[lonelyminer (Peter Šurda)]

Demand deposits ARE available for exchange.
I have $100 in a USD bank.  I realize I need to pay a bill. I withdraw $10.  The $100 was available for exchange.

In your example, it's the reserves, rather than the deposits, that are available for exchange. The withdrawal was a redemption, not an exchange. The exchange occurs when you trade with a person other than the issuer, and you need to withdraw prior to that.

[DeathAndTaxes]

Your first modifications is dubious. If someone someone doesn't think BTC on deposit has the same value then ... they won't deposit.

Value is subjective. Merely because a depositor thinks that the deposit has value, it does not follow that other people think that too. If they are not a customer of the same issuer, they might just as well see this derivative instrument as an additional cost.

It doesn't matter what others think.  If/when they get paid it will be in BTC.

What are you talking about customer of the same issuer?  You would be paid in BTC and nothing but BTC.

I deposit 100 BTC in to Bitcoin International Bank.
Obviously by my actions I value that demand deposit = 100 BTC in a private key I control.

I owe you 10 BTC.
I am not going to have BIB issue a BIB-share for 10 BTC and hand that to you.

I owe you 10 BTC
I withdraw 10 BTC from BIB and transfer it to you.
You receive 10 "genuine and authentic" BTC.

Your trust in BIB is immaterial.  The amount of trust collectively in BIB determines how much deposits BIB attracts.
No trust/value/utility = no deposits.
High trust/value/utility = high deposits.

Where does everyone seem to have this false impression that fractional reserves involves derivatives, alternate scripts, demand notes, etc?

Think back the last time someone paid you with USD and had USD on demand at a local bank.  Did they
a) withdraw USD and pay you with USD (via CC, debitcard, cash from ATM, ACH, etc).
OR
b) have their local bank print of bogus-localbank script and pay you with that which you then had to evaluate on its value, deposit at your local bank and request they contact the other bank and arrange repayment in USD
?

Why does anyone think that Bitcoin banking would involve B when every fractional reserve banking system in the world involves A?

[lonelyminer (Peter Šurda)]

I can imagine scenarios involving a ripple-like debt network where p2p, short term lending would be far more common…you would have more adhoc, informal lending on a p2p basis where trust is more important than formal risk assessment…and you would also have your more traditional lending where risk assessment is rigorous.  Under such circumstances, it's conceivable that the multiplier could actually be greater.

I am not sure about this interpretation. Why should decentralisation decrease the likelihood of redemption?