We have had a lively discussion of usury here, and the question has come up as to just how it is that fractional reserve banking (FRB for short, hereafter) creates new money. Or, “creates new money”, if you would prefer. Part of the problem with that discussion is whether we should call what is generated 'money' or not. Before we delve into that, I think it would be beneficial to lay out in great detail just what we understand by the term FRB. Here is a short, pithy, fairly direct explanation of the skeleton of the animal:
How Fractional Reserve Banking Works
When you put your money into a bank, the bank is required to keep a certain percentage, a fraction, of that money on reserve at the bank, but the bank can lend the rest out. For instance, if you deposit $100,000 at the bank and the bank has a reserve requirement of 10 percent, the bank must keep $10,000 of your money on reserve and can lend out the $90,000.
In essence, the bank has taken $100,000 and has turned it into $190,000 by giving you a $100,000 credit on your deposits and then lending the additional $90,000 out to someone else.
Now, if you take this out a little further, you will see that your original $100,000 can become $1,000,000 by the time it is all over. Here’s how:
- You deposit $100,000 Your bank loans someone else $90,000
- That person deposits $90,000 Their bank loans someone else $81,000
- That person deposits $81,000 Their bank loans someone else $72,900
- That person deposits $72,900 Their bank loans someone else $65,610
- That person deposits $65,610 Their bank loans someone else $59,049
- That person deposits $59,049 Their bank loans someone else $53,144
- That person deposits $53,144 Their bank loans someone else $47,829
- And so on
Ultimately, your initial $100,000 can grow into $1,000,000 with a 10 percent reserve requirement.
To find out exactly how much money the fractional reserve banking system can theoretically create with your initial deposit, you can use the Money Multiplier equation:
– Total Money Created = Initial Deposit x (1 / Reserve Requirement)
For example, with the numbers we have used above, you equation would look like this:
– $1,000,000 = $100,000 x (1 / 0.10)
I would like to flesh out this skeleton with what I understand of what else is going on, so that we can see just what the argument really is all about. I will admit that I am no expert in this stuff, and if others have pertinent additions, I will be glad of it.
First, I am going to take us back to the good ol’ days when money was, generally, gold coins, silver, copper, and sometimes iron coins. There probably never was a time when ALL money was that, there were times and places when other items were used as a medium of exchange, but by the year 1500 most of it was metal coin. (And when I say “coin” I mean ONLY physical coins that you could hold in your hand and test for gold alloy.)
Harkening back to a prior post: Mr. Goldsmith had to have a store of gold on hand to do his smithing work – jewelry work and gold leaf, perhaps along with making coins. Hence he had to have a good locking vault, and maybe a guard or three. Other people had higher risks involved with their holding on to gold, so if they were very worried, they might bring their gold to Goldsmith to store it for them. (It might be gold coin, but it could also be other gold – jewelry, lamps, a cross, etc.) He would (if sensible) charge a fee for the service. And he would give them a receipt which testified to (a) he had received their gold for storage, and (b) they could redeem the gold by presenting his note. The receipt was both proof of the gold deposited and a contract for its return.
By and by with growing trade and wealth, a lot of gold and silver in the form of coin was stored this way. And people would sometimes, instead of going to Goldsmith to get their coin out of storage and buy some big-ticket item, simply draft a note to Goldsmith and hand it to Fred the seller that said “please pay Fred 10 gold coins out of my deposit”, i.e. a draft on the buyer’s account with Goldsmith. Fred could go and do the redeeming for the buyer, Bob, when he actually wanted the gold if he didn’t want to carry the gold off right away.
However, Fred also had an account with Goldsmith, and didn’t want to carry the gold away, and so when he would present the check in Bob’s name, Goldsmith would simply make a change in his accounts book, adding 10 to Fred’s account and subtracting 10 from Bob’s.
But then the question comes up: how much did Fred trust Bob, as to whether he really had the full 10 gold coins on deposit? Sure, Bob had the original receipt on hand that said Goldsmith had received 10 coins on deposit, but maybe Bob had already made checks for some (or all) of the money. And how would that all work if Bob went to another city where nobody knew him and that his tailoring business was “good for it”? Eventually, Goldsmith wrote out generic, unspecified “banker’s notes”, in small (or, sometimes, large) denominations, that simply meant “good for redeeming X gold coins”. These were something like bearer’s bonds, effectively, which Bob could hand to Fred as evidence of the gold on deposit at Goldsmith’s. Whether Fred would accept the note depended less on whether he trusted Bob and more on whether he trusted Goldsmith to have what he claimed in the note: gold on deposit available for redemption, that he would honor the note, etc.
Now let’s go back to the FRB scenario above, and fold it in. John goes out on a trading voyage that is highly successful, and comes back with 10,000 gold coins, which he certainly does not want to store. He takes the money to Goldsmith, and instead of a fully written out receipt for his money he receives 1000 banker’s notes each for 10 gold coins. He accepts the notes as under the same standard as the former receipts: he recognizes that there are risks of putting his money in Goldsmith’s hands, there could be a fire or a theft. Goldsmith could take the money and run. These are the risks he accepts. The notes are a proof of deposit and a contract in the same way the receipt was. However, the notes have fine print on the back that specify a “10 day hold period” that Goldsmith can require between when the note is presented for redemption and when Goldsmith has to fulfill the claim. (Goldsmith explains that he stores larger deposits at an off-site secure location, and it takes time to retrieve them.)
Able comes to Goldsmith and says: I want to buy a house worth 10,000, I have 1,000, but I need to borrow 9,000. Goldsmith offers to hand him 9,000 gold coins as a loan. But Able doesn’t want to carry that much, it’s too heavy and too risky. So Goldsmith hands him a letter of credit testifying to the 9,000 ready on account for Able, and has Able sign a contract. According to the contract, Able must (a) pay off the debt in 10 years; and (b) if Goldsmith asks for the full debt to be repaid, Able has 9 days to cough up the cash: he can get alternate loan funding from another banker, or sell the house, or whatever. (This clause (b) is, of course, just in case John asks for his 10,000 gold coins back – Goldsmith never expects to have to invoke this clause. See * below for a discussion of alternatives to this clause (b).) Naturally Able has to pay interest, and has to initial clause (c) of the contract that provides the house becomes collateral for the debt.
Able buys the house from Baker, by handing over the letter of credit and 1,000 in coin. Baker goes to Goldsmith and has him “remove” the 9,000 from Able’s account and post it to his own account.
Now Baker has been hankering to start a deli, and asks Goldsmith for a loan. He will use his 1,000 coin, 9,000 from his account, and borrow another 8,100 from Goldsmith (call him G). He will convert the bottom floor of his town home to the purpose, and will buy machinery, shelving, and food stock with the money. G has him sign a contract requiring (a) that he repay the loan over 10 years, and (b) if G asks for the money, he has 8 days to cough up the coin. (G explains that he never expects to invoke (b), but it is there “just in case”.) And of course, (c) he has to sign over a subordinate collateral assignment of his home to G to the extent of 1/3 of it (the bottom floor), since another lender is the primary assignee for the home as collateral. G provides Baker with 900 ten-coin (new) bankers notes for the “money” that is “in” his account (the 9,000 that came from Able), and a letter of credit for the other 8,100 that G is lending him.
Now Baker buys all his items from Charlie, the restaurant supplier in town. He provides his 9,000 in banker’s notes and the 8,100 in credit, along with the 1,000 coin. Charlie takes the notes and letter and has G credit HIS account with the 8,100 in credit from Baker.
Charlie has hankered all along to turn his back 40 acres into a truck farm, to provide his own food stuff to sell to restaurants. But he needed an expensive tractor, which he now has money to buy. He gets a letter of credit from G for 7,290 and new banker’s notes for his 8,100 on account. Charlie plans to buy the tractor (plus a storage barn) for 24,390, which is equal to his 17,100 in banker’s notes and 7,290 in credit. G has Charlie sign a contract providing (a) that Charlie has to repay the loan over 10 years, and (b) if G asks for the money, he has 7 days to cough up the coin. (G explains that he naturally never expects to invoke (b), but it is there just in case John comes along asking for his coin.) And, of course, (c), he makes over a primary assignment of collateral on the tractor, and a subordinate assignment on the 40 acre farm.
Charlie buys the tractor from the local John Deere dealer, Dan, and hires Dan’s outfit to build him the barn. Now Dan has 17,100 in banker’s notes, and 7,290 in credit from G. He goes to G to convert the letter of credit and has him credit _his_ account with 7,290.
Dan all along has wanted to expand his showroom so he can show off his latest and best equipment, and now he has the “money”, he has 24,390 available: 17,100 in banker’s notes and 7,290 on account with G. He wants to hire Echo to do the construction, to the tune of 30,951. He get’s a loan from G (i.e. a letter of credit) for 6,561. He naturally has to sign over a contact with the terms (a) that he will repay the loan in 5 years, and (b) if G asks for the money back he has 6 days to cough up the cold gold coin. Dan also pulls out 7,290 from his account, in banker’s notes. And of course, he signs over a subordinate assignment in the dealership’s physical building (supposedly worth 40,000, but he owes Tom 33,000 on it).
Dan pays Echo the 24,390 in banker’s notes and the 7,290 letter of credit.
And so on. At each stage, the seller of the prior stage takes the letter of credit and replaces it for banker’s notes. G started with 10,000 coins, and so far he has written out 24,390 in banker’s notes on them, plus the first 10,000 in notes given to John. (And he isn’t done yet, it will keep going till the total is 100,000, if he keeps to his 10% reserve ratio.)
The parties involved have been using the banker’s notes and letters of credit “as if” they were the gold coin they represent, but of course what they are really is a _promise_ with respect to gold coin. They are, therefore, a risk: that G is not lying about having the gold coin, that he won’t take his gold and skip town, that G’s house won’t burn down or be broken into. That, at least, is the risk that John knowingly accepted in taking HIS banker’s notes as receipt for the coin deposited. (This is aside from the risk, inherent, that the gold itself may change in value. The notes are promises as to being redeemable for gold coin, not promises that the gold will remain valuable.)
But notice, critically, that in Echo’s hands, those banker’s notes now have a heck of a different cast of risk than they did at the start. In Baker’s hands, the notes depended on Able paying his loan, or Able being able to get alternate loan funding, or selling his house within 9 days. In Charlie’s hands they depended on Able and also on Baker paying his loan, or Baker being able to get alternate funding, or being able to sell something worth 8,100 in 8 days…etc down to Dan and then Echo. If Echo tells G that he wants the gold coin, G has 10,000 coins on hand, but has to call in the other notes to get the rest of the 30,951 he claims from G.
Note, please, that this model of FRB does not inherently involve usury. All the loans were with collateral. We can simply assume (since nothing else was written into the contracts) that G’s reach for repayment is limited to the collateral, not to the person of the borrowers. Nor does the FRB inherently involve “fiat money.” At every step of the way, the lending was in terms of gold coins that G actually has on hand, or could have on hand by calling in loans. Nor does the paper "money" inherently mean "sovereign money" - i.e. an obligation of government, at every step the paper items being exchanged were originated by Goldsmith who paid a clerk to make out fancy, difficult-to-copy "notes" that were contracts on an obligation by G. This just clean, pure, unadulterated FRB without other frills like regulatory constraints.
*[ In theory, G could have managed item (b) of his contracts differently. He could have written his banker’s notes with explicit conditions. His notes to Charlie could have read “if Able and Baker do not repay their loans with coin upon recall, this note is redeemable instead for a proportionate share of my loan contract with them rather than in coin”. And his notes to Dan could have said “if Able, Baker, and Charlie, do not repay their loans with coin upon recall, this note is redeemable instead for a proportionate share of ‘my contract with Able, Baker, Charlie’ rather than in coin“. This would be a slightly different arrangement, and would create an explicit difference between the notes given to John and the notes give to Echo. In practice, though, nobody would ever make such conditions an explicit element of banker’s notes, because they would have to alter their notes for each change in total accounts, which is completely impractical.
The other alternative to (b) is for G to never mention clause (b) at all, and not have any fine print on the back of the notes. I would call this an outright fraud perpetrated upon those who receive the notes, though, because G has an obligation to respect the possibility of some of the parties somewhere along the line redeeming the notes, and ignoring that obligation is effectively just handing out promises with no intention of fulfilling them. Eventually G has to assume that John will ask for the gold back. If there is only a possibility of the other parties redeeming a portion of the notes, depending on all the other parties not doing so, then the character of the contract promise inherent in the notes (including John’s) changes: “redeemable for gold coin as long as you get there early enough in the bank run”. A note that means (but does not say) “redeemable for gold coin…but not REALLY if you actually, you, know, ask for the coin, only if you notionally ask for it” isn’t a real contract promise – i.e it’s fraud. This ties in with a later point. Hence my having G use clause (b) is an attempt to avoid purely fraudulent bank notes or purely fraudulent FRB, and to SHOW the real relationships being created.]
Three things to note that are important, I think: First, after this program has been going on for a while, John probably would (or should) know up front what G was going to do with the gold when he put it on deposit, in part because G would be paying John for the privilege of handling the money, rather than John paying G for the storage costs. But when G FIRST STARTED doing this, there could be no presumption on John’s part that he was getting those secondary risk elements for his deposit, and if the contract with G did not explicitly state “G may lend out my coin” then John would not have been buying into that scheme of risk. There’s a big difference between “I am storing my goods with you and accept the risk of damage or theft” and “I am depositing my goods with you for investment and I accept the risk of some of the investments turning bad.” In the scenario, the FIRST time G starts to get into lending, John didn’t deposit the gold coin for further investing.
If John wants to invest in other people’s ventures, he can indeed hire G to do it for him or do it himself. I happen to think that it is a lot more prudent, looking at the view from 30,000 over all of society, for John to do at least some of his own investing, because he is then more likely to have a real understanding of the risks he is buying, instead of a hands-off (and heads-off) attitude of “let someone else have the headaches.” What is patently unreasonable is for G to subject John’s deposit to investment risks that he never had reason to expect because it was a new scheme and not part of the written contract.
Secondly, in theory, G can indeed fulfill a demand by Echo for the full 30,951 in coin if every loan call is successful. But there’s the catch, that conditional statement “If…”. Suppose all of the other borrowers who had other debt (and who had primary collateral on the home, farm, and dealership with someone else) had their loans with one single lender, say Silversmith. And suppose that between them these two effectively hold all the loans in town. You could easily have a situation where a single demand for redemption by Zed (the last guy in line) could not only force Goldsmith and Silversmith to attempt to call all the loans in town, but that there simply are not 100,000 coins in town to redeem the calls. (In fact, if loans constitute enough of the town economy, a call on ALL of the loans in town would always fail, because of the interlocking nature of the businesses.) In practice – even without a call on all loans affecting the entire town - not every loan call is going to be successful. Some of the property won’t be worth what debt they cover. Some of the property cannot be sold in time.
Much more importantly, there is in principle nothing that prevents G from lending out without “adequate” collateral, which does in fact happen if the borrower’s reputation for making money (or business model) is convincing even though there is no sufficient hard object of collateral. My uncle got business loans by convincing the bank he had a great business model without sufficient up front collateral that obviously would succeed in paying off the debt (the increase to the value of his business by expanding would only be WORTH the money put into the plant (and salable for that same amount on the open market) it if the business proposition was correct. If incorrect, nobody else would pay that kind of good money for the building.) Even if G calls in all the loans with collateral, he cannot expect to always make good on loans calls. Or, to put it another way, G’s “balance sheet” holds so-called “assets” that are measured in terms of the notional value of the loans, not the ACTUAL value of the collateral. G does not have to, for example, do a recall on loans whose collateral has dropped in value to restore his target reserve ratio. The reserve ratio is a preferential notion IN HIS HEAD, not a physical or structural limit to the possibility of lending. The standard in the industry has changed over time, with the rate now at about 3% (so I have read.)
In essence, the risks associated with FRB based on notes reflecting gold (and silver, copper, …) coin include not only the risks of each individual loan being paid off, but ALSO the risk that calls for actual, physical coin are “manageable.” Inherently, one of the critical features of managing that risk is making sure your reserve-to-lending ratio is not too low. And one of the risks is a misjudgment about how likely (and large) demands for redemption will occur in relation to the available REAL (as opposed to on paper) options each borrower has when he tries to comply. The risks to this judgment / misjudgment are not obvious. For one thing, it is almost impossible to correctly “measure” the probable rate of calls, demands for pay-off on the loans, withdrawal of cash from deposit accounts, for when “things go rotten.” A system that only works when everything is going right is a system that is inherently unstable. For another thing, since the reserve ratio is a mental CHOICE by the lender, not an inherent given to the system, there is nothing that prevents a banker to decide to change his ratio (unless he promised to keep a certain ratio to his depositors, which normally does NOT happen. Show me a bank account that explicitly states a reserve ratio.) So if investors jump in with deposits under a 10% reserve standard, and their banker decides he likes more risk and changes to 5% reserve, they are suddenly taking more risk than they planned.
Thirdly, all of this picture hangs prettily on notes that refer to a specific physical item of value. It is true that the market value of those coins of gold can (and did) change over time and market fluctuations. But there was a specific, physical CONTENT to the promises that could be pointed to as a base, rooting value of the notes – yes, taking into account the risks implicit in a note versus the gold coin in hand. The value the notes point to at any time is the value of physical gold coins. The value of the notes themselves is the combination of (a) the value of the coins themselves, and (b) the risks inherent in leaving them in play versus physically collecting them in cash. That (b) is somewhat subjectively ascertained. There is something objective that the subjective user’s implicit use hangs on, there is an objective referent to the notes so the notes are a promise ABOUT SOMETHING ELSE. Yes, accepting the promise is, inherently, accepting a risk. That risk is rooted in “what will people choose to value gold (or whatever the external referent is) at?”
When you switch to a system for which there is no root referent for the notes (i.e. fiat money), AND maintain the fractional reserve lending, I believe that the success of the system may depend critically on not paying attention to the inter-related web of relationships of risk, with the conjoined lack of a root referent. The risks of the inter-related loans are then saddled with a new risk: “what will people chose to value the other notes in the system for” against which there is no external referent.
There is a qualitative difference between a system of dollar notes that reflect an external root referent and a system of dollar notes that don’t. I believe that the nature of the latter is insufficiently examined. It is a relatively new experiment, virtually untested in any real sense (the most recent period of history, with the US moving away from the gold standard, is a drop in the bucket in terms of a historical frame of experience). I suspect, also, that our cycles of bubble and burst are, at least with a certain amount of probability, indicators of underlying problems that are not fully recognized.