Skip to content

Ep. 48 Bob Explains the Highlights of His New QJAE Article on Fractional Reserve Banking

Bob goes solo to explain some of the most important points in his new QJAE article on the fractional reserve banking debate. Bob shows why Mises thought *any* issuance of fiduciary media caused the boom-bust cycle, and he points out a major flaw in George Selgin’s defense of fractional reserve banking.
Mentioned in the Episode and Other Links of Interest:

The audio production for this episode was provided by Podsworth Media.

About the author, Robert

Christian and economist, Research Assistant Professor with the Free Market Institute at Texas Tech, Senior Fellow with the Mises Institute, and co-host with Tom Woods of the podcast "Contra Krugman."


  1. Ty Fyter on 08/08/2019 at 2:29 AM


    i had a question that is related to this (tangentially); i’m doing an undergrad econ finance (IFIM) course and on the topic of commercial banking (presumably FRB), my lecturer has *strongly* stated that loans are made first and then deposits are made second. This seems to run opposite to my understanding of FRB where deposits are made of which some is loaned out to others.

    A corollary statement he made was that in order for anyone to save money, money has to be borrowed first.

    How am I supposed to think about these claims? (particularly the 1st one about loans and deposits)

    Any feedback would be greatly clarifying,

    • Manan Sharma on 08/10/2019 at 4:54 PM

      Your teacher is right. Today currency is created by federal reserve by giving loans to banks. Usa has debt based currency. But your teacher is underestimating the impact of inflation in destroying the savings.

      Proof: since 1987 when federal reserve started printing money the ratio of savings to gdp of America has gone down.

    • Dusan Vilicic on 08/11/2019 at 3:08 PM

      My guess would be tah he is referring to the fact that under the current system, for bank notes to be created, the bank first has to borrow from the Central Bank (the Federal Reserve in the US). Under the current monetary system’s logic, at first there are no Fed notes. The banks go to the CB and take out a loan. That money is “printed” (electronically created, currently) by the CB. Then the bank can loan to the rest of people (or buy government bonds) that new money. Once the money has been lent, the lendees can deposit it in the banks. So in this current system it’s like your teacher said. As far as I understand, at least.

    • fiatjaf on 08/13/2019 at 1:16 PM

      If you remove the first deposit from the historical table of events then yeah, it will look like banks first loan out then get deposits.

      Now answering your question, your lecturer is coming from the perspective that says “all money is created by debt” according to which without an initial loan there would be no money to be deposited. This is a fundamental principle of the MMT (Modern Monetary Theory, don’t get misled by this imposing name, they’re a narrow group of crazy people) proponents Bob has talked about extensively both in this show and elsewhere.

      These claims make some sense but fall apart entirely if you look at the history of money and banking.

    • Martin Brock on 08/13/2019 at 3:28 PM

      Credit is not lending money. Credit is paying over time in installments. Think of it as a rent to own agreement. I want to sell my house for $240,000, but no one interested in my house has $240k lying around, so I agree to sell you the house for $240k payable over 20 years in 240 monthly installments of $1000. In addition to the $1000, you must also pay me rent monthly, but since you gradually accumulate equity in the house, your rent payments gradually decline. When you’ve bought half the house, you pay half the initial rent and so on.

      We can also arrange a payment schedule wherein your monthly payments are fixed, so the equity you purchase each month increases as your rent payments decline, but let’s stick with the fixed equity payments for the sake of simplicity. To account for this transaction, I create 240 share certificates. Each month, when you pay the $1000+rent, I give you one share of the house. Your monthly rent is a fixed amount, say $6, per share you don’t yet possess. After the first month, I sell you one share for $1000 and receive $1200 in rent. After the second month, I possess 239 shares and receive $2195 ($1000 for a share and 239*5$ for rent). After the last month, I receive $1005, and thereafter, you own the house in full.

      Clearly, these shares are valuable to me, and they’re as valuable to anyone who’d like to receive the same payments, so I can bargain with them. Bob has a car he’d like to sell for $10,000, so I offer to exchange ten of my shares for the car, and he accepts. For the next ten months, you pay Bob $1000/month, and you also pay him $5/share he possesses while paying me $5 for each share I still possess. After ten months, you continue making equity and rent payments to me directly.

      A bank is a firm accounting for transactions of this kind. The shares function as money, but this money exists only after you and I agree to your purchase of the house in installments. I outsource the accounting to the bank. The bank issues the shares, gives them to me, and I deposit them in the bank. You make payments to the bank, and the bank makes payments to my account, gradually replacing my shares with dollars (base money), but in the meantime, since the shares are valuable and negotiable, I may use them as money as well.

      • Martin Brock on 08/13/2019 at 11:24 PM

        Correction: At first, I write “$6 per share” but intended $5 per share. $5/month/share amounts to $60/year/share, and since each share is worth $1000, the effective, annual interest rate is six percent.

  2. LP on 08/08/2019 at 4:57 PM

    With your last point, it is worth noting that, if the goal is a 10% increase in real savings, neither issuing loans to everyone nor mining new gold would accomplish that. In both cases, if everyone suddenly had 10% more in their bank account, we would expect prices to rise about 10% overnight, which would leave purchasing power unchanged. The only way to avoid that is to introduce the new currency only at particular places, which then lets the first receivers to increase their purchasing power at the expence of last receivers. Even there, the average real savings does not increase.

    • Robert Murphy on 08/09/2019 at 2:32 PM

      No LP this isn’t right. Yes, other things equal, if 10% more gold coins suddenly were dropped from the sky, then we’d expect prices to rise by 10%. But that’s because (by construction in this example) we are assuming the demand to hold money was the same.

      In contrast, if everyone in the community wants to increase the purchasing power of their gold holdings by 10%, then if the miners dug up that amount of extra gold and it found its way into everyone’s cash balances, there would be no reason for prices to rise. People wouldn’t be spending more per day at the original price structure; they would add the new gold coins to have permanently larger cash balances.

      Put it another way: If there were NO new gold coins coming into the market, and people had a 10% increase in the demand to hold gold coins, then prices would have to fall. So if, instead, new gold coins flow in to satisfy the extra demand, that keeps prices from falling; it doesn’t raise them. (Or if you want: Yes, the new gold coins raise prices *relative to the counterfactual in which prices otherwise would have fallen*.)

      • LP on 08/10/2019 at 6:39 PM

        I see what you’re saying. It seems odd because if the desire to hold cash went back to its previous level, we would expect prices to rise. Then again, it’s already the case that if the desire to hold cash decreases, we would expect prices to rise. And I think this is where I went wrong, yet still have a point.

        Presumably, the desire to increase cash balances is not ex nihilo, rather it is in response to some suspected future event (natural disaster, war, et cetera). Each individual wants to increase the purchasing power of their savings by 10% so that, if the future calamity happens, they will better be able to purchase the resources they need to endure it. My point is that, excepting the direct utility of the mined gold, and especially in the case of fiat currency, the real resources of the society have not increased. So when the calamity strikes, all the additional cash savings will do is cause the scarcity prices to spike higher than they would have. Moreover, if the cash balance increase is due to FRB, without prices rising, the people increasing their cash holdings may not realize that the prices will go up so far in a crisis.

        In concrete terms, if there’s a small village, say 200 people, on a floodplain, and there’s a boatwright with 10 row boats for sale, the people in the village may decide they need to save enough to be able to buy a boat if it looks like the river is about to flood. If they all increase their cash on hand, but don’t actually get the boatwright to make additional rowboats, there won’t be enough on hand when the river floods.

  3. Geoff on 08/09/2019 at 3:29 PM

    The most interesting aspect of this episode for me is about halfway through when Bob explains that he is still in favor of free banking, in spite of his belief that fractional reserve banking of any kind contributes to the business cycle. This seems to have also been Mises’ position.

    This is the one issue that I thought perhaps there could be a contradiction between Austrian school economics and libertarianism, but after listening to this episode, I may have changed my mind. Austrian economics, at least to many, seem to point to fractional reserve banking as contributing to the boom/ bust cycle. But libertarianism tells us that we should not initiate force. However, just because something may not seem optimal, it doesn’t mean we need aggression to “correct” it. Free banking, in which market forces tend to limit FRB, is optimal in the sense that any state intervention will make things far less optimal.

    I think a good analogy can made with money. Gold may not seem to be the optimal form of money. Having a form of money that never changes in supply may be better, but this doesn’t exist. It is actually optimal to allow the market to choose our form of money, which historically has been gold. We could put the state in charge of money and say that the money supply should never change, but we know this will never happen. In addition, it is a form of coercion.

    One thing that frequently frustrates me with the FRB debate within the Austro-libertarian community is that we are typically presented with just two sides. One side is Selgin/ White who defend FRB in economic terms. The other side comes from Rothbard/ Salerno/ Block who say that FRB is fraud and should be illegal. (I am not 100% sure if this is Salerno’s position, but he doesn’t emphasize that FRB should still be legal if he believes so.) The crazy thing is that it is often self-identified anarcho-capitalists who want to make FRB illegal. If there is no state, by whom will it be made illegal? And whose head are you going to point the gun at to prevent FRB?

    Those looking to make FRB illegal say it is fraud, but how would it be fraud if all parties agree to what is happening? They say that you can’t have multiple claims at the same time, but this is the whole basis of the insurance industry, which I don’t think they think should be illegal. If everyone got into a car accident tomorrow, the insurance companies would not be able to make good on their promises.

    I believe Bob’s position is the correct one from an Austro-libertarian standpoint. I would just change the emphasis. The primary point is that we should have a system of free banking (the free market). Market forces would severely restrict the ability of banks to engage in FRB. The secondary point is that, to any degree that FRB still took place, it could have small impacts in causing a boom/ bust cycle, although these impacts may not even be noticeable in the size of our economy.

    • Robert Murphy on 08/09/2019 at 5:41 PM

      Thanks for the comments. I agree that sometimes it is not clear to me what a given 100% reserve theorist wants the government in the world to do right now. But, in fairness, I think that is a secondary issue. By the same token, I definitely think murder would be illegal in Ancapistan, but does that mean I want the State right now to levy property taxes to fund police departments and courts that arrest murderers? Not so clear.

    • LP on 08/10/2019 at 5:55 PM

      I think the first note to make is that, in the absence of central banks and the FDIC, the amount of fractional reserve banking would probably be quite limited (it was historically at least). While this is still expansionary, much as new gold mining is expansionary, the severity of the cycles is likely to be much less severe, and much shorter duration.

      Additionally, Joseph Schumpeter has some work on the particular nature of business cycles, and how mild and short ones aren’t a bad thing. While his theory on their causes is somewhat lacking, his theory on business cycles allowing for innovation is good. So it’s not clear that the ‘optimal’ state is one with perfectly fixed amounts of credit, and I would posit it’s certainly the case that the harm caused by business cycles (when not inflamed by the state) is less than the harm caused by states trying to prevent them (even if the state in question did so by prohibiting FRB).

    • Dusan Vilicic on 08/11/2019 at 4:47 PM


      There is a key difference between a deposit and an insurance. A deposit contract is, in essence, a safe-keeping contract. I give you something of mine for you to store and keep safe. An insurance contract is very different: I pay you for you to give me money in case some agreed upon thing ‘X’ happens. You are *definitely not* telling me you’ll give me back the money I gave you. In fact, you’re not saying that you’ll do anything at all with the money I give you. In this sense, insurance is closer to a betting contract than to a deposit contract. You simply promise you will pay me an amount if ‘X’ happens. You might be Scrooge McDuck and have a huge fortune to draw upon, and you might be charging me a merely symbolic amount, clearly not enough to turn a profit. That does not matter there.

      In the deposit, the money is still my money. I simply gave it to you to store it. If you give my money to someone else, then you’re breaching the contract. When the thing I gave you is fungible, the contract may state that you will store the amount, but not the same molecules. Even then, if I store 10 ounces of gold of a given quality, you still have to keep the amount, otherwise you’d be breaching the contract. This is easier seen if you have only one client: me. This way, the “law of large numbers” does not get in the way of analysis. If you lend 9 out of the 10 ounces I gave you, you’d be breaching the contract, clearly.

      Also, you claim that in an anarcho-capitalist society it would not be possible to have that kind of law. But law is not state-imposed, necessarily. Private law is a thing, and it has existed in the past too, 100% privately-provided law. So the fact that an ancap says “X should be illigal” does not automatically constitute a contradiction. Especially in a setup like the one imagined by David Friedman, a polycentric one. There you can have agencies that offer libertarian NAP-based law, agencies that offer Sharia Law, others that offer theonomist law, illegal FRB, legal FRB, whatever. Even in other setups, you can simply say that a contract that is impossible to satisfy is null and void from the start, and FRB being based on those (two separate property claims to the same object), it would be illegitimate and unenforceable.

      In any case, the issue is not so obvious, clear-cut and simply decided as you seem to portray, is my main point.

      • RS on 08/15/2019 at 5:10 AM

        @Geoff and @Dusan: I like the comparison with insurance, since it shows the flexibility contracts may have. If the demand deposit at a bank is not subject to 100% reserve, I assume that the bank doesn’t tell me that I have a “guarantee” to withdraw my cash anytime. If they did, I could see the point of calling it fraud.

        More likely, the bank will say when I open a demand deposit, that they give me a ledger that shows how much I “deposited”, and that I can use it to exchange with other people. And that I can “demand” at any time to withdraw my deposit, but that there is no guarantee that I actually get it. The bank might be closed (incl. bank holiday), the ATM may be empty, they might not have $100k in cash in the vault, the borrowers might not have made their payments or defaulted, in which case I have to wait until some of the money is recovered. Or I was lazy, and showed up last when there was a bank run. In any case, there might be a long line, and they might only assign one slow teller giving out deposits. Or they might limit the withdrawal per day.

        It is obvious to me that these are the terms I accept if I open and use a checking account these days. With the wrinkle, that FDIC and money printing could bail out a bank in my favor (in the latter case devaluing my money, however).

        So in summary, the free market is still the best answer, where people could demand a 100% reserve requirement or even the form of deposits (e.g. gold). But this is either not in great demand right now, or the government interferes too much with those business models, or both. (Clearly, the government doesn’t like you using gold and puts capital-gains taxes on it, and it won’t accept it in any transactions either).

  4. Greg on 08/11/2019 at 5:35 PM


    They do have 100% reserve banking, it’s called a safety deposit box. How can a bank pay you interest if it doesn’t loan money? How can a bank loan money if it has 100% reserves? Am I missing something here?

    To me there are two factors that contribute to unstable banking:
    1. Long term loans versus short term or uncertain deposits
    2. Greedy banksters who loan money to bad prospects knowing that Uncle Sam will bail them out when things go awry.

    The solution to the 2nd issue is easy. Get government out of banking including the FDIC insurance scam.

    The solution to the 1st seems manageable with some level of reserve and interbank lending.

    No wonder economic used to be considered a part of philosophy. According to George Berkley, “Philosophy is the profession where you kick up a lot of dust then complain you can’t see.”

    • Robert Murphy on 08/13/2019 at 3:48 PM


      A safety deposit box isn’t the same as 100% reserve banking. If I go to the grocery store, I can’t swipe a debit card that tells the bank, “Open up Bob’s box and take out $42.80 and put it in Wal-Mart’s box a few feet to the right in the vault.”

      And right, a bank can’t pay interest on, or lend out, reserves that are tied to checking accounts, in a 100% reserve system. But banks can still be credit intermediaries, for example by selling CDs. Or, they can have genuine time-deposit savings accounts, where you deposit your money and then can’t touch it for (say) a year.

  5. Dwain Dibley on 08/12/2019 at 8:11 PM

    This will probably fall on closed minds but…
    Banks generate credited deposit account entries as loans. Banks also generate credited deposit account entries as payments. Deposit accounts are a record of the amounts of ‘money proper’ owed to the deposit account holders by the banks. You do not spend your record of deposit, you spend the money that the record of deposit represents, even when the banks do not have that money.

    Fractional reserve banking is a banking system in which only a fraction of bank deposits are backed by actual cash on hand and are available for withdrawal. 100% of all deposit accounts (demand and savings) are records of bank debt to deposit account holders, they are not money in any sense.

    Unless the banks are issuing checks in the amount of, banks do not issue fiduciary media. A deposit account is not a fiduciary media. Debiting or crediting your record of deposit with the amounts is not a fiduciary media. You are not exchanging anything with anyone when you use a debit card, you are simply transferring your obligation to pay to the bank. In that same vein; you are not redeeming or converting anything when you withdraw your money from a bank. The bank did not give you anything when you deposited your money, and you are not giving them anything when you withdraw your money. The bank adjusting your record of deposit to reflect the transaction is not an exchange.

    • Martin Brock on 08/14/2019 at 2:06 PM

      Credit money is not backed by reserves. It’s backed by collateral. Banks hold reserves to satisfy their depositors’ demand for cash (base money or gold under a gold standard). Their deposits and circulating notes have value, because the collateral securing their credit (like mortgaged real estate) has value. This value has nothing fundamentally do with the value of cash in a vault.

      Without a central bank, FDIC and the like, when a bank runs low on reserves, it must call loans (require the sale of a house for cash immediately rather than over the course of the loan for example). Banking contracts had stipulations entitling banks to call loans and to require depositors to wait a period before withdrawing deposits. Jimmy Stewart explains it all during a bank run in “It’s a Wonderful Life”.

  6. Giovanni on 08/13/2019 at 2:08 PM

    In many years studying Austrian economics, some confusions about “the” ABCT (the quotes are because I perceive different theories depending on who is explaining them and in which context, although they share the same fundamentals) keep coming to my mind in diverse forms. Upon listening to this episode, I was able to note the questions that appeared, and they are the following:

    1. If a new influx of genuine gold could trigger a boom-bust cycle, why couldn’t couldn’t the same happen with an increase in genuine savings? For example, if people abstain from consuming shoes and instead go barefoot and save, that will decrease interest rates and may prompt malinvestiment in the umbrella industry. The fact that people saved on shoes has no relation at all with some expected desire to consume more umbrellas in the future.

    2. Don’t we have multiple interest rates in the economy? (I mean money rates: the rate from which Joe borrows from a bank isn’t the same Mary does, and these are different from what one bank pays to another and so on.) Isn’t talking about “the interest rate” an aggregation Keynesian sin?

    3. Is it really required for the new inflow of money to go to the general loan market for a boom-bust cycle to be triggered? What if the new inflow of money goes to the housing market only — for example, the Fed could make policies that use newly-created money to lower the interest rates _for house purchases only_. That would surely trigger a boom in the construction sector, and from that sector the boom would spread to other related businesses and adjacent sectors, and since the construction sector is so big that would have enormous impacts.

    4. Rethinking 1 and 3, why do these effects have to come from an increase in the quantity of money? For example, if the government taxes the shoe sector a lot and use the money to buy and subsidize umbrella purchases, wouldn’t that cause a ton of malinvestment in the umbrella sector? And wouldn’t these malinvestments spread to other adjacent businesses and sectors?

    Sure, this situation could not go forever without new money entering the economy, so it’s not sustainable. Other sectors in the economy would have to shrink while these are growing and maybe other details would be different from a boom created by new money. But hey, isn’t the classic boom from the ABCT also “unsustainable”?

    Maybe I should try to raise my questions in the secret listeners group.

  7. Transformer on 08/14/2019 at 9:00 PM

    I very much enjoyed the paper and have a couple of comments on section III.

    1. In the example you give of a boy who saves fiat dollars and then pays his fiat dollar into a FRB bank account: Assuming this is analogous to a boy saving a gold coin and then paying that coin into a FRB bank account then from the perspective of the FRB system I do no think the paying in of the gold would count as an act of savings but rather one of new reserves being added to the system. Free Banking theorists would see it as causing a (temporary) monetary disequilibrium. A mature FRB system is I think assumed to have a stable quantity of outside money

    2. In the example you give of an economy where everyone wants to increase cash balances by 10% then I suppose if you also assume everyone is identical in their borrowing requirements then the outcome you seem to find absurd (everyone increase their cash balances by in effect borrowing from themselves) would indeed be the equilibrium outcome. This is not an absurd result though. People build up the cash balances by decreasing consumption and FRBs lend out these balances by reducing interest rates to induce borrowers who by assumption have to be the same group of identical people. In a world where everyone is not assumed to be identical there might still be a tendency for many people to simultaneously wish to increase cash balances – but the resultant increased lending will go to those with relatively high time preference or businesses who will use them to increase the production of “future goods”.

  8. Enrico on 08/17/2019 at 11:25 PM

    From 00:58:00 to 1:03:00, you state that banks have to increse lending (“issuing more money via the loan market”) in order to let people increase their cash holdings. The following argument is that the interest rate drops and that the *supposed* Selgin’s explanation (“everybody lent himself the money”) is not convincing.
    However, the first premise doesn’t seem correct to me. Suppose that I deposit $100 in gold coins into a bank account. In a free-banking system, if I suddenly want to increase my cash holdings, I can go to the bank and withdraw (let’s say) $10 in cash – i.e. banknotes issued by my bank. This act doesn’t increase lending at all – no money is issued via the loan market – but surely my cash holdings are increasing. Therefore, I’m just swapping $10 from my bank account for $10 in banknotes; in other words, I’m exchanging a type of bank liability for another type. To repeat, there is no increase in lending, hence no decrease in the interest rate.
    In a free-banking system, cash holdings can increase by having people withdrawing banknotes from their checking acconts; of course, their accounts are then debited the corresponding amount of money. Thus, I think that Bob is misinterpreting how a free-banking system would work – at least, according to the free-bankers.

  9. Enrico on 08/18/2019 at 12:08 AM

    @Bob (Part 2)
    From 00:51:20 onward, you state that depositing money into a checking account doesn’t constitute saving money. But you also said (00:46:30) that banks deposits are equivalent to very-short-term loans to the bank, and later (00:52:19) you stated that lending money to the bank (i.e. via a certificate of deposit, CD) is an act of saving. So, the two things together mean that depositing money into a bank account is an act of saving, because it’s like having a very-short-term CD renewed over time.
    – If I open a $100-worth 1-year CD, than sell it for $100.1 after 1 month, and buy $100-worth stuff…nobody denies that I saved $100 for 1 month.
    – If I deposit $100 into a checking accont, then withdraw $100 after 1 month, and buy $100-worth stuff, would anybody deny that I saved $100 for 1 month? What’s the practical difference with the previous line?
    One issue to be clarified is what constitutes saving and consuming. If I held $100 in my pocket for 1 month, I’m saving because I’m consuming no present good, i.e. I’m restraining from consumption. I’ll consume a present good only when I’ll buy one. Having money in my hand constitutes no consumption.

    • Robert Murphy on 08/19/2019 at 6:36 PM


      Something is screwy with my browser and I can’t play the episode right now to check, but I think you misunderstood me when you wrote: “But you also said (00:46:30) that banks deposits are equivalent to very-short-term loans to the bank”

      No, *I* don’t think that, it’s the FRFB people who think that.

      • Enrico on 09/04/2019 at 3:48 PM

        Thanks for the clarification, and sorry for my misunderstanding. Still, I don’t really see any practical difference between very-short-term loans and bank deposits…

  10. Enrico on 08/19/2019 at 2:53 PM

    Part 3
    What does Selgin have in mind when he talks about “people wanting to increase their cash holdings”? He is referring to a scenario where people decrease their spending (i.e. they are saving more than before) and keep the money under their mattresses instead of depositing it into a bank (*).
    If banks cannot issue banknotes, there is a problem: people are hoarding gold coins under their mattresses, so their increased savings cannot be “translated” into more investments. If banks can issue banknotes, instead, there is no problem: people can hoard banknotes, while gold coins stay inside bank vaults; since people are spending less than before, there are less money withdrawals from bank accounts, so banks can increase lending without the risk of losing too much gold. In the end: more savings -> more loans.

    (*) Of course, no problem would arise if the savings were deposited into the banking system, because then banks could issue more loans due to the increased amount of savings.

  11. […] the critics will wonder how banking could even exist with 100% reserves. (For example, here is my recent podcast episode on the topic, and in the comments you can see that some listeners were mystified by the […]

  12. […] Bob Murphy Show episode 48, in which Bob discusses his new QJAE paper on the FRB […]

Leave a Comment