Actually, the bank records as an asset the IOU of Joe Smith, in the amount of say $10000. And records as a liability the checking account of Joe Smith, $10000 (more than the yesterday). The bank's assets and liabilities have both increased by the same amount. Joe Smith's assets and liabilities have both increased by the same amount. So far, no "cash" (physical banknotes) has entered a particle of it.
Then Joe Smith writes a check to his supplier. This is an IOU of Joe Smith, underwritten by the bank. You can tell because it has the underwriter's routing number at the bottom. It can be pay to bearer or pay to an addressee. Joe Smith now owes in two places at once - the loan and the check - while being owed in one - the bank to him for the checking account. In return he acquires some other asset, presumably, from his supplier.
The supplier used to own some other asset, now he has traded it for a check from Joe Smith drawn on Joe's bank.
Pause. Check not deposited yet. What has happened?
The supplier could have just lent Joe Smith the value of the goods he sold him. Joe could have paid him that way. His loan IOU would be with the supplier instead of with the bank. The bank wouldn't enter the picture at all. There would be one extra contract with possible future value in existence, Joe's IOU. Real assets in existence are of course unchanging, but their *value* may not be unchanged, at all. They have moved in ownership and use, and Joe may have better things to do with them than the supplier. Presumably, urgent enough better things that he offered say 7% interest on his IOU.
If the moved goods are enough more useful in Joe's hands than in the supplier's hands to repay that interest, then net value can readily have been gained by all concerned, without a single new physical item coming into existence. Since all items are themselves just arrangements, and goods are valuable arrangements of scarce items, this should be completely unsurprising. But acres of misconception about the real origins of value mislead people about it.
If Joe is *wrong* about the usefulness of the item to him, compared to what he offered for it, 7% interest per year, then the action may have destroyed value overall. And he will bear the loss, because he owns the asset and all it can do for him, but owes the 7% per year. If Joe is *right* about the usefuless of the item to him, then net value has indeed been created, and he will reap it. If in the middling case he offered the whole extra value of the item to him in interest, then the whole society and he himself and his lender, will all break even. And note well, the debt can be fully repaid.
Now back to the intermediated case. The supplier didn't loan directly to Joe, because the supplier has no knowledge or opinion about the soundness of Joe's plans, and no idea whether Joe can pay 7% interest and will return the principle. The supplier doesn't want to become an expert in such things to sell his item, which he specializes in being good at producing effeciently, not assessing ultimate users of it, etc. So he takes a check instead.
Why does he like taking a check? Is he going to put it in his dresser drawer for a rainy day? No, he likes it because it is denominated in a common unit of account, and drawn on a reputable payer, bank. In the supplier's experience, when bank says somebody can pay, they can. He finds his own bankers readily accept checks drawn on Joe's bank and will pay cash for them, or a bank deposit, at full face value.
So the supplier takes the check to his bank and "deposits" it. This really means he has sold the check to his bank and immediately loaned his bank the proceeds. Let's look at that transaction.
Before it, the supplier owns a check. After it, he owns a bank deposit. His bank, before the transaction, has nothing to do with any of it. After it, it holds the check and owes the supplier a $10000 bank balance.
Now the supplier's bank goes to clearing with Joe's bank. After they net off all their other transactions, they find (compared to what would have happened without this one) that $10000 more has been drawn on Joe's bank than on his supplier's. Joe's bank reduces Joe's checking account by the amount of the check, and owes supplier's bank $10000. Which in practice Joe's bank borrows from anyone in creation at 3%.
Joe owns an asset but has an IOU outstanding against it paying 7% interest. Joe's bank owns his IOU but has borrowed the same value from anyone in creation at 3%. His supplier has a bank balance at supplier's bank. Anyone anywhere has loaned Joe's bank the same amount at lower interest.
First order, it is very likely the supplier. His proceeds are sitting in his money market account, and are relent by his bank into the LIBOR market, where Joe's bank borrowers them.
In fact, the extra bank balance created when Joe and his bank agreed on the loan, can careen around as much as you like but can't be destroyed, until Joe repays the loan or Joe's bank writes it off. It can end on deposit in Joe's bank, or some other one, but it isn't going to disappear. It is, in fact, now part of the money supply.
Yes, a liability was also thereby created, and someone has to fund it. But it would be funded if Joe never drew on the account. It would be funded if his supplier left the proceeds in a money market. It would be funded if after thirty eight intermediate steps it ends up back in Joe's bank as a retiree's CD held by Joe's father in law. Wherever is stays or goes, somebody is going to own it, and if it remains a bank deposit, it is going to be lendable to fund Joe's loan, directly or indirectly.
But maybe someone instead wants physical federal reserve notes. That's fine, their holding those still funds the banking system - even cost less than CDs. The "earning" bank is just the Fed in that case. Holding an FRN is completely indistinguishable from having a checking account deposit of the same amount, with the Fed as your bank.
OK, but what about the question of the soundness of the loan, under the intermediated case? Has it changed? It has not. If Joe's father in law lent him the money at 3% and Joe paid 3% to him and 4% to his favorite charity, the same soundness questions would arise. The bank earns because it is more useful at getting this stuff to happen than Joe's unassisted efforts, let alone his favorite charity. Really, 1% of that 7 is an insurance premium because Joe may fail. And another 2% of it is a running cost of the bank. Leaving it a slender but positive profit for its intermediation.
Is it inevitable that Joe will invest in worthless assets, empty buildings, stupid schemes that all fail, saddling himself and everyone else with real welfare losses? Nope, not at all. He can gain or lose, depends entirely on how sound his bets were. "He should be forbidden to do so because when he fails, he hurts the rest of us, not just himself". That principle will leave no economic freedom of any kind, anywhere. Anything you do economically, that fails, hurts others as well as yourself. You use resources that could have been used better elsewhere.
Suppose Joe instead first scrimps and saves the full value of his purchase from his supplier, out of income. Does this eliminate the risk of loss from his purchase? Not at all. It means the bar his use of the item has to clear is his real time preference in scrimping and saving, instead of the 7% offered by the bank. It means he is invested about a third in his father in law's CD, now held by him, about a third in the bank's stock, and about a third in his speculative venture.
That may or may not be a sounder portfolio. It may or may not have a higher return than being all in his speculative venture. But it can still fail, by exactly the same process as any other entrepenurial command, that resources should cease to be held by party A and purpose A, and should instead shift to party B and purpose B. It can also fail even if its return is positive nominally, if it fails to clear the bar of Joe's own real time preference (then it fails him), or the whole society's (meaning his scrimping and saving didn't even benefit others, since it reduced current incomes too much for anything it gained longer term).
All of the above would be true without any fractional reserve system at all, without any bank regulation at all. In practice, whenever banks are not directly up against their maximum reserves, and even when they are if the Fed is being accomodative about adding extra reserves because they are, the above is the actual process of money creation.
The Fed simply controls the total *demand deposits* - which means *only* checking accounts, not savings or CDs - that all the banks together create for their customers, to a multiple of the size of the Fed's own balance sheet. The Fed itself increases the size or decreases the size of its balance sheet, with just as much freedom as an unregulated version of Joe's bank. In return for its promise to provide FRNs on demand, for any demand deposit, it requires banks to hold a fixed amount of physical FRNs in their vaults or deposits at Fed branch banks.
Notice, this means there was a reserve requirement at the instant of creation of the checking account for Joe, but only a small fraction of its face, meetable by bank vault cash or Fed deposits. But also notice, once the money thereby created transfers to a CD or savings account at the other end of the intermediation chain, there is no reserve requirement against it, at the Fed or in vault cash. The Fed only directly controls M1, not the other money measures.
The limits on the growth of the other money measures is solely the prudence of the banks and the Basel capital standards, which apply to all of their assets and liabilities (though risk weighted in quite generous ways), not just that subset of their liabilities that are customer checking accounts.
Also notice that the public has a free choice between the reserved-against transaction forms, physical notes or checking accounts, and the non-reserved forms, savings and CDs. If the public wants to hold more money as a mere safe form of investment, then its demand for the second sort will rise, while its demand for the first may remain unchanged or fall.
The whole system is *designed* to not be restrictive about such desires. A surge in merely safety seeking, not spending power seeking, desire to hold money, will not meet any reserve requirement "bar" preventing quantity from moving. This is deliberate - it is meant to prevent such panic surges in money demand from changing the exchange *value* of money. (Quantity moves instead).
Now examine recent matters knowing all this. You will see M1 hasn't moved since March of 2005. Since prices have, the real narrow money supply is actually contracting. The Fed started that deliberately to break the real estate bubble and did so - too many Joe's were clearly making clearly bad bets, and being accomodated. The Fed deliberately raised the hurdle any such brainstorm had to clear, to be funded.
In the last 2 years, on the other hand, broader money has increased without narrow money increasing. Why? Because fewer supplier pay-ees are using Joe's loan proceeds to buy more raw material or hire more workers, and more of them are using them to instead run up their safety-seeking balances of CDs and institutional money market accounts etc. Flight to safety demand to *hold* non-transactions bank-money, is not a sign of too-rapid money creation - quite the reverse.
If you see M1 accelerating relative to MZM with prices not (yet) moving, then the Fed is driving up narrow money trying to increase the nominal value of transactions. If you see MZM decelerating and M1 catching it, as prices rise strongly, then the public is trying to get out of dollar balances into real goods. But if you see M1 flat and MZM rising, it means the public is trying to get into dollar balances, and out of riskier assets. Which is deflationary (if money quantity didn't move, money value would rise, aka all prices would fall), not inflationary.
the federal reserve will not clear your bank's check for Joe's car against your banks funds based upon Joe's IOU. They will only do so on the basis of actual reserves held on deposit with the federal reserve by deducting the $10,000 from the same. Your bank actually has to have real money on deposit. Even Toddster agreed with that point.