Fiduciary Cycles

Blogger “Lord Keynes” criticizes the use of the term “funny money” as being purposefully loaded.  I agree, but I suggested separating Austrians between those who believe that fractional reserve banking is fraud and those who do not.  It strikes me that the size of the latter group is growing, and the former dwindling (at the very least, more and more people are simply switching to supporting “free banking,” whether it be with fractional or full reserves).1  In response, LK writes the argument I would have least expected him to use.  Paraphrased,

  • Why would a fractional reserve free banking system not lead to recurring industrial cycles;
  • According to LK, the “whole point” of Mises’ and Hayek’s work in the area was to show how fiduciary expansion leads to business cycles;
  • In fact, how do promissory notes, bills of exchange, and checks not cause industrial cycles.

My purpose here is not to inflame another debate between full reservists and fractional reservists, but just to defend the White/Selgin (also referred to W/S here) model of free banking against LK’s claims.  In fact, “defend” might be the wrong term, since what this really boils down to is an illustration of just how poorly LK understands both the Austrian theory of industrial fluctuations and the theory of free banking.

Allow me to begin at the middle: Mises’ and Hayek’s theory of severe intertemporal discoordination.  Their ideas have to be understood in proper context and it also has to be accepted that they made mistakes — mistakes that do not invalidate the broad argument, but rather imperfections.  The theory of intertemporal discoordination (which I have discussed at length elsewhere: 6 Jan. 2011, 26 July 2011, and 12 August 2011) can be summed up as arguing that growth in the supply of money will lead to the creation of malinvestment throughout the structure of production, if this growth takes place in the “arteries” (my own term) that channel savings from one person to another — as long as the growth is over and above the supply of “real savings.”  Understanding this is crucial: what both Mises and Hayek were discussing is a distortion in the pricing process that makes capital goods seem more abundant and available than they actually are.

The basic idea behind malinvestment and the business cycle seems to be lost on a lot of people, and not just non-Austrians.  We are not talking about a lack of consumer demand for the final product, because consumer output prices will adjust to reflect their marginal valuation.  We are talking about the inability to complete investments, because the quantity demanded of capital goods outstrips the supply.  When these “artificial savings” are no longer available, entrepreneurs can no longer afford to bid away capital goods which have risen in price as a result of the monetary expansion.  This is why Austrians refer to the idea of investment over and beyond the stock of real savings.

In Hayek’s early writing (I have in mind his 1933 [1929] article “Monetary Theory and the Trade Cycle;” specifically, chapter four), he does actually believe that fractional reserve banking leads to recurrent business cycles.  He thought that industrial fluctuations were an unfortunate side effect of a more efficient use of savings. Hayek’s beliefs with regards to the debate, I think, mature between 1929 and the 1970s (i.e. “Choice in Currency” [1976] and “The Denationalization of Money” [1976]).  I think the Hayek of 1929 is correct to argue that the Austrian theory is an endogenous one, but I do not think he realizes what factors cause this recurring endogenous problem to be possible. That is, there are important institutional characteristics — one obvious one being the Federal Reserve — that makes endogenous business cycles possible.  This is why, in later life, Hayek opted to argue that the banking system should be reformed, by means of full privatization and deregulation.

Hayek calls Mises’ version of the theory an exogenous one, but I think this is a mistake that originates from the lack of insight on the institutional characteristics of the banking system.  Mises recognized that the problem is one of cartelization and regulation.  Also, contra LK, Mises never opposed the issuance of fiduciary media.  In Human Action [1949], for instance, Mises calls fiduciary media an important tool, but highlights crucial limitations to a bank’s ability to extend credit (pp. 431–441).  The mechanism by which this occurs is actually roughly similar to that in the White/Selgin model: excess notes are spent and ultimately returned to the bank for redemption, threatening a bank’s stability.  Why do banks not seem to be affected by these supposed limits in most, if not all, modern banking systems?  Government intervention.

I am not saying Mises was a fractional reserve free banker, although I do not think Mises was a full reserve absolutist, either (however, Mises did argue in favor of full reserves in lieu of free banking, to constrain the capabilities of a cartelized and regulated banking system).  A pretty convincing and authoritative comparison between Mises and W/S is provided by Salerno’s May 2010 Mises Daily piece (although, his calling W/S free bankers part of a “Neo Banking School” seems to me a bit too broad).

Two main points that should be taken from what I have written so far,

  1. Austrian business cycles are caused by “artificial savings;”
  2. Neither Hayek, except very early on, nor Mises thought fractional reserve banking is the source of the problem — they thought it was fractional reserve banking by an institution which characteristics and abilities have been distorted by government intervention.

How does the White/Selgin model of fractional reserve banking fit in all of this?  In Selgin’s The Theory of Free Banking (a book that, like most economic treatises, needs to be read from cover to cover), two general means of reserve reduction are presented.  Quoting directly from my article “Prices and the Demand for Money,” these are,

  1. Over time, “inside money” — banknotes (money substitute) — will replace “outside money” — the original commodity money — as the predominate form of currency in circulation. As the demand for outside money falls and the demand for inside money rises, banks will be given the opportunity to shed unnecessary reserves of commodity money. In other words, the less bank clients demand outside money, the less outside money a bank actually has to hold;
  2. A rise in the demand to hold inside money will lead to a reduction in the volume of banknotes in circulation, in turn leading to a reduction of the volume of banknotes returning to issuing banks. This gives the issuing banks an opportunity to issue more fiduciary media. Inversely, when the demand for money falls, banks must reduce the quantity of banknotes issued (by, for example, having a loan repaid and not reissuing that money substitute).

The S/W model, therefore, assumes that none of these two phenomena will lead to a dramatic2 over-issue of fiduciary media.  This is because, in principle, the supply of money in circulation3 is stable.  In the case of (1) bank notes originally act as money-certificates, replacing outside money in circulation.  The act of turning some of those money-certificates into fiduciary media by selling part of a bank’s reserve stock of outside money does not change the fundamental fact that (1) alone maintains the supply of money stable.  The description for (2) above explains the mechanism, but I will reiterate: the idea is that money held (not spent) is no longer circulating, so it can be replaced by a new banknote (or equivalent unit of credit or what have you).  What (2) does, though, (and this is the most controversial part, I think, within Austrian circles) is distribute this held purchasing power to entrepreneurs, shifting spending to the non-consumer stages of production.  The W/S rationale is that holding money represents savings, because the individual holding money is deferring from present consumption.

This is my attempted reconciliation.  Mises, as I have already said, was not a fractional reserve free banker in the same vein as W/S, et. al.  He did not recognize any of the above two mechanisms.  Mises is clear, though, that individual banks which over-issue will risk their stability, because when over-issued fiduciary media return to banks for redemption, banks will find their reserves too scarce.  Again, this is the same mechanism present in the W/S model.  The important difference is that free bankers who prefer Mises’ theoretical exposition tend to think that fractional reserve lending will be minimized by “market forces” and free bankers of the W/S kind — such as myself — tend to think that reason (1) behind fractional reserves will exert the most pressure in reducing the size of reserves of outside money and (2) will cause subtle fluctuations in the stock of fiduciary media.  Whatever occurs, it is easy to see how free banking would not lead to recurrent industrial fluctuations, and any intertemporal discoordination that does occur will not be to the same degree as it occurs in cartelized/regulated banking system (with a government currency monopoly).


1.  What this suggests is that the size of the population of well-read Austrians who really believe that fiduciary media is “funny money” is actually substantially low and dwindling.  Those who use it in the media are either: (a) part of the small group of Austrians who are full reserve absolutists or (b) using the term precisely because it holds a negative connotation.  An alternative reason could be that some of the people who use the term “funny money” are thinking specifically of fiduciary media created that would not have been created in an alternative free market banking system — in other words, it is a fuzzy term that has a nebulous meaning to encompass the entirety of the banking cartel.  I actually suspect that it this last reason that explains most of the usages of the term, with a little bit of (b) mixed in as well (every Austrian knows “funny money” is not an academic term, but their intentions are not academic).

2.  Remember, these theories are meant to be ideal types; they explain the workings of a system by abstracting from certain details.  In this case, the abstraction is disequilibrium.  This is why I write “dramatic over-issue,” since over-issuance (and under-issuance) will occur in a world characterized by disequilibrium and decentralized knowledge.  It is also conceivable that some banks might fail due to over-issuance — entrepreneurs can make mistakes.

3.  The idea of “in circulation” is key.  Mises 1998 (1949), “The only vehicle of credit expansion is circulation credit,” p. 431.

37 thoughts on “Fiduciary Cycles

  1. Gene Callahan

    “We are not talking about a lack of consumer demand for the final product, because consumer output prices will adjust to reflect their marginal valuation.”

    Jonathan, have you read Sowell’s book on Say’s Law? He notes that in the case of Malthus and Sismondi, he shows that they recognized that prices would so adjust, but that they might do so at a level that caused losses, and produced a general slump in production.

    1. Jonathan Finegold Catalán Post author

      I haven’t read Sowell’s book, but I agree that Malthus and Sismondi are right. I think it’s accurate to assume that a major part of the problem is that credit expansion will increase the price of capital goods (marginal costs). It also strikes me that the largest drop in demand will be in the demand for capital goods, since the reversal of the boom should bring about a leftward movement of respective demand curves (as a result of malinvestment and a lack of new, “cheap” credit).

      There’s a difference, though, between a leftward shift in the demand curves of different non-consumer stages of production and a leftward shift in the demand curve of consumers due to an increase in idle money. The former has to do with a skewing of the process of price imputation from marginal valuations and the latter has to do with a change in marginal valuations.

    2. Björn

      IMHO economists of all kinds underestimate these kinds of losses. In our private lives we’re used to thinking of economic value as governed by some law of conservation, sort of like energy. In entrepreneurial life you quickly learn that huge losses and gains are the default.

      Take this extreme example of malinvestment driven by credit expansion: It cost $14 billion to build, and now it will most likely just sink back into the sea. There is no “output price that reflects marginal valuation”.

      1. Jonathan Finegold Catalán Post author

        That is a great example of an unfinished project (and a capital good). However, it’s price does reflect marginal valuation (in this case imputed from the price of related consumer goods, which as we can see are not very highly valued) — this is manifested in its idleness.

    1. Jonathan Finegold Catalán Post author

      (Speak of the devil! I’ve actually started to go over and read the Critical Review debate between you, Horwitz and Garrison in the 1990s.)

      I’m not sure if you want me to refer to the essay of yours you link to. I’ll provide some mild responses, but will focus mostly on what you emphasize here (expectations). Actually, I think the “expectations argument” is the least powerful criticism (the most powerful being anything that seriously challenged the Austrian theory of price formation, whether that of economic goods or its theory of interest [on this note, my intention is to read Maclachlan’s critical comparison of the liquidity preference theory and the pure time preference theory sometime soon] — although, I will preemptively suggest that Sraffa’s 1932 critique is not as important as some think it), largely because it doesn’t consider the factors that individual entrepreneurs actually have to account for. This is more fundamental than the rejection of rational or fully informed expectations, and would be true even if all entrepreneurs were fully informed of potential costs.

      One of the first economists (that I know of) to suggest considering the effects of informed expectations was actually Ludwig Lachmann, in his 1943 paper “The Role of Expectations in Economics as a Social Science.” In a short comment (“Elastic Expectations and the Austrian Theory of the Trade Cycle”), Mises argues that his theory of the business cycle actually do consider changing expectations. But, I think that Mises’ response is weak.

      The creation of an unsustainable widening and lengthening of the structure of production does not lead to a 100% probability of loss. In fact, it can lead to a lot of profit; one can go as far as to say that if an entrepreneur perfectly predicted both the boom and bust that entrepreneur would become quite wealthy, because he could take advantage of the rising prices of producers’ goods and later accurately speculate on the drop in prices and general economic collapse. In any case, if we disaggregate the population of entrepreneurs and categorize them within the different stages of production they decide to invest in, we see that their respective “aggregate supply functions” will look very different. This is because entrepreneurs of the third stage and beyond aren’t interested in the outcome of production of the second stage; their income comes from the demand originating from second stage and beyond investment. And, as long as credit expansion continues to accelerate or more-or-less maintain itself, as long as these entrepreneurs make good choices they will profit.

      This is good evidence that favors the idea that credit expansion will induce a widening and lengthening of the structure of production. The extent of bust-induced failure depends on which entrepreneurs are better at predicting the downturn — and, contra the claims of some Austrians, it’s not all about looking at Fed policy (for instance, there could be some event which triggers deflation before the Fed decides to sufficiently slow or completely cease credit expansion — the accelerating failures to repay subprime and ARM loans could possibly be one example). Some entrepreneurs did manage to cut losses or even turn a profit — some entrepreneurs with a general knowledge of Austrian theory come to mind (Peter Schiff?) — in this most recent recession. Most entrepreneurs, though, didn’t see it coming.

      For these reasons, I just have never really been convinced by the expectations argument.

      1. Greg Hill

        Jonathan, you write, “One can go as far as to say that if an entrepreneur perfectly predicted both the boom and bust that entrepreneur would become quite wealthy.” Granted, it would be very difficult to “perfectly” predict “both the boom and bust.” But if the Austrian theory of the business cycle is superior to its alternatives, wouldn’t you expect Goldman Sachs, Boeing, John Deere, etc., to hire Austrian economists to do their forecasting?

        Three questions regarding the “structure of production”:

        1. Do you think firms typically have a list of alternative techniques that vary mainly in terms of how long it takes to produce something?
        2. If a firm is choosing between a project that takes 3 years to complete and one that takes 10 years to complete, doesn’t uncertainty about cost, demand, etc., trump interest rate considerations? and
        3. Have you looked at any of the research on the interest rate elasticity of investment?

        1. Jonathan Finegold Catalán Post author

          But if the Austrian theory of the business cycle is superior to its alternatives, wouldn’t you expect Goldman Sachs, Boeing, John Deere, etc., to hire Austrian economists to do their forecasting?

          If any heterodox, whether Austrian or post-Keynesian, critique of mainstream modeling is right, then why is it that mainstream modeling is still more popular? Why are there so many people who believe in exclusive faiths? These types of questions don’t really have anything to do with how “superior” or accurate any given theory is.

          As for your questions,

          1 & 2. This is why the re-switching argument is irrelevant. Austrian business cycle theory has nothing to do with how long production techniques will be for individual firms. Actually, Austrians will tell you that the firm will choose the cheapest, shortest production technique. Lengthening and widening refers to,

          Widening: An increase in the number of firms in a particular stage of production (or an increase in production of goods of that stage);
          Lengthening: An increase in the number of stages of production.

          3. No, I haven’t, although I suspect that it varies. Right now, for instance, I assume that a fall in the rate of interest would induce only a relatively minor increase in investment.

  2. Chris

    Can a a free banker answer a few honest questions for me?

    (1) During the transition period from 100% reserves/commodity money to fractional reserves obviously the amount of reserves held by the bank will fall. How is this any different from a central bank announcing a reduction in the reserve requirement? If the Fed arbitrarily announced it was cutting reserve requirements from 10% to 5%, surely Austrians could agree this would create malivestment. How is it any different when the free banking system reduces reserves from 100% to say, 20%? Won’t this cause malinvestment that will need to be liquidated (and a contraction of fiduciary media in the process)? It seems to me the attempted to transition from 100% reserves to fractional reserves will get stuck in a perpetual cycle of boom and bust (and inflation and deflation). Feel free to correct me, but doesn’t Selgin’s Theory of Free Banking just gloss over the transition period and jump straight to the alleged inter-workings of a “mature” free banking system?

    (2) Under fractional reserve free banking how is an addition to the monetary base due to gold mining any different from when a central bank arbitrarily creates bank reserves at the click of a mouse? Again, Austrians would say the latter would create malinvestment as the banks proceed to pyramid money and credit on top of the new reserves, artificially reducing interest rates. Does gold mining not do the exact same thing, that is it expands the base on top of which banks can pyramid new credit? Why would this not produce malivestment and destabilize the free banking system?

    (3) The Ricardian theory of the business cycle suggested that when one country inflates faster than another, that there would be an outflow of reserves from the inflating country to the country that inflates at a slower rate, ultimately causing bank runs and a contraction of the money supply in the country that originally inflated. Rothbard argued that the same thing happened between individual banks. That is, if one bank inflates faster than all others that it will lose reserves to its non-inflating competitors, essentially causing a run or at least a contraction of credit in the inflating bank. Assuming the validity of this theory, how can the banking system transition from 100% reserves to fractional reserves without cooperation/collusion on the part of the banks? White and Selgin rule out the possibility of cooperation (correctly, I believe) on the grounds that such cooperation is unstable on the free market. How then does the entire banking system move from 100% reserves to fractional reserves without cooperation?

    (4) Has anyone ever read Hulsmann’s “Has Fractional Reserve Banking Really Past the Market Test?”

    If so are there any responses to the points he makes? Specifically that baring confusion (possibly due to a lack of proper disclosure) in the minds of the public between 100% reserve notes and fractional reserve notes, how could it be that financial instruments denominated in gold, bearing credit risk and liquidity risk (ie. FR notes), could out compete gold and 100% money certificates to become the generally accepted medium of exchange? That implies that FR notes (again with risk) are more liquid than the very commodity they are denominated in. How could that be? This seems to contradict Mengarian theory of money where the most salable of all commodities out-competes all others.

    I put this question to Steve Horowitz and he just brushed it off by saying the empirical evidence proves that FR notes have historical out-competed gold. But just because the empirical data contradicts a theory doesn’t mean the theory is false. There could be some sort of intervention driving the discrepancy. For example, the presence of housing shortages in NY doesn’t on its own refute the theory that prices adjust upwards in response to shortages. If Hulsmann’s theory that FR notes only out-competed gold historically only because inadequate disclosure caused people to believe the notes were as good as gold is correct, then this would seem to blow up the theory of free banking.

    White wrote a response to this paper,
    Though, I found his arguments unconvincing. Basically he argued that since he personally cannot figure out a non-clunky way of charging for storage on 100% reserve notes, that this must be the reason that FR notes out-competed gold and money certificates. Nevermind that Larry White is an economist not an entrepreneur. Academia is full of economists who make similar claims for public goods. That since they, from the comfort of their offices, cannot figure out how to provide for a non-rivalrous, non-excludable good that must mean neither can any entrepreneur.

    Any answers to these questions?

    1. Jonathan Finegold Catalán Post author

      1. Reserve requirements don’t dictate the volume of outstanding credit. Bank lending is constrained by the risk of default; i.e. the ability to meet outstanding obligations (and bank notes are liabilities to the bank).

      2. This could be equally as applicable to any banking system, as long as the new money enters circulation through arteries that reach entrepreneurs. In any case, someone can conceivably take gold coin to a bank and ask for a money certificate (a bank note, for all intents and purposes). This doesn’t allow a bank to suddenly pyramid fiduciary media on top of this gold. The bank note represents a liability, since it allows redemption of the deposit at demand. As I said in response to (1), a bank’s ability to extend credit depends on its financial stability, which is a measure of its ability to pay liabilities.

      3. This implicitly assumes that fractional reserve free banking (FRFB) is inflationary. I don’t intend to argue about this, since this argument has been made countless times before. But, the rationale behind the rejection of the idea that FRFB is inflationary is in the post you’re replying to.

      4. It doesn’t “outcompete” gold, since the bearer of the note can redeem the note for the original deposit of outside money if he so wanted. But, the viability of FRFB lies in that the bearers of bank notes trust in the bank’s stability. But, to argue that FRFB is a priori unstable is, to again, entirely weave around the dispute of whether it is or not. In any case, White’s argument makes sense. Assume the costs of storing money in a bank with full reserves is the storage fee, which we’ll call F. Alternatively, the costs associated with FRFB bank notes are p(L), where L is whatever loss is associated with holding a FRFB note and p is the probability of this loss occurring. If p(L) > F, the person will prefer a money-certificate; if p(L) < F, the person will prefer the bank note. FR Banks will tend not to charge on deposits since they earn profit by using demand deposits as possible funds to loan out. If p(L) > F, a FR bank might compete with full reserve banks by paying interest on demand deposits, where the new cost is p(L) – interest earnings.

  3. Chris

    Thanks for the reply Johnathan.

    Let me just press these issues a little more.

    1) I recognize that bank lending is constrained by risk of default, but the assumption free bankers make is that the level of reserves (given the risk of adverse clearing) will be something less than 100%. If true then the transition from 100% reserves to fractional reserves will be inflationary (it won’t be covered by an increase in money demand.) Only once the bank reaches its self imposed limit (again given the risk of adverse clearing) that inflation will cease.

    2) under a 100% reserve system there is no reason to expect newly mined gold to enter the economy exclusively through the loan markets. The gold miners will pay their employees and suppliers with the new gold, presumably the new gold would be spent in rough proportion to the current societal consumption/investment ratio leaving interest rates unchanged. The critique I made seems to only apply to a FROM system where the new gold adds to the monetary base thus allowing for artificial credit expansion.

  4. Chris

    I’m typing this only my phone so it just accidentally posted that last comment before I was finished.

    Additionally, you said that more gold mining doesn’t imply banks will automatically expand credit because whether it expands or not will be based on its ability to pay. But that ability to pay has increased if it sees its gold reserves increasing, which then makes it profitable to expand credit (unbacked by an increase in money demand).

    3) I wasn’t arguing that FR banking is inflationary in the sense that an increase in credit that IS backed my an increase in monetary demand amounts to inflation. I concede to the free bankers that such an increase would be stable. But I’m again talking about the transition from 100% reserves down to the lowest level possible given the risk of adverse clearing. Such a credit expansion does NOT coincide with an increase in the demand for money.

    4) Fair enough. But it seems hard to accept the argument that “everyone knew” back then that FR notes were really financial instruments and not claims to fixed sums of money. Just looking atsome historical notes, they make no mention of credit risk or risk of suspension. They just said “payable on demand”. I’m not sure that would be enough for the financially illiterate to realize what they were accepting. The only way to find that out would be to do it and see if people do accept them as money or not.

  5. Chris

    On a similar note. It seems that free bankers tend to exaggerate the cost of storage. charges .0018% per year for storage. So if I had an average of say $1,000 in a checking account (which is about what I keep), then my monthly storage fee would be about 15 cents. Which is negligible. That 15 cents/month would need to outweigh the risk of my bank suspending withdraws on my FR checking account to get me to accept fiduciary media. For me, I’m not sure it would.

  6. Chris

    One final point. Suppose a creative entrepreneur decided to offer “free storage” to its customers as a means of attracting people to his other banking products. Or suppose he charges advertisers a fee to advertise on the front or back of his notes the uses this revenue to subsidize the storage fees then proceeds to offer free storage. Now p(L) > F. Our entrepreneur would put his FR competitors out of business. Or at least out of the business of issuing fiduciary media (people still may use FR accounts for investment purposes).

    Despite all the work done by scholars I still feel like this theory was hastily thought through.

  7. Jonathan Finegold Catalán Post author

    1. I’m not sure how you come to this conclusion. Let’s say period 1 is all money-certificates, such that 1,000 units of gold are covered by 1,000 money-certificates. Overtime, banks shed 200 units of gold. The money supply in circulation is still 1,000 bank notes, 800 of which are still money-certificates and 200 of which are fiduciary media. This is not inflationary.

    2. All new outside money that adds itself to the stock of money is an increase in the money base, since the money base is defined by outside money. As I’ve already explained — the exchange of gold for bank notes increases bank liabilities. This constrains the ability of banks to issue a greater number of bak notes… so no, the addition of new reserves of outside money does not automatically allow banks to “pyramid” fiduciary media on them.

    3. Yes, but again, this doesn’t actually affect the amount of money in circulation.

    4. I don’t understand what you’re trying to say. FR notes are physically the same as money-certificates. What differentiates them is that they are objectively unbacked by outside money. The circulation of money-substitutes is a gradual process — nobody has ever made the argument otherwise (see the various literature on the evolution of a free banking system).

    The cost of storage would only be outweighed by the risk of loss if the risk of loss by fractional reserves is high enough. If a FRFB system is stable, then this perceived risk of loss shouldn’t be high just because it’s a fractional reserve bank. The stability of these banks would be assured over time, with these types of banks gradually replacing full reserve banks because (a) they are inherently more profitable per client and (b) a sufficient amount of clients now are indifferent between money-certificates and fiduciary media. And, no, free storage wouldn’t be enough to out-compete fractional reserve banking. In my example, F can be 0 and [P(L) – interest] can be negative, which translates to a net benefit.

  8. Chris

    1) Johnathan, I don’t want to accuse you of not understanding how a fractional-reserve system works, but your example here is wrong. You say that if we start with 1000 fully backed money certificates and then the back sheds 200 units, that the money supply will not expand (800 fully backed notes plus 200 fiduciary media). But this is wrong. The original 1000 notes continue to circulate and then the bank loans out an additional 200 (either 200 gold coins from the vault or 200 newly printed notes). The money supply increases to 1200 units (1000 original notes plus the 200 new notes). This is text book FR banking. White and Selgin gloss over this transition period and proceed to analyze a “mature” FR system. Some people have complained about the redistribution surrounding the inflation during the transition period and they brush it off by saying that its just the profit to the banks for devoloping a superior banking system. Ok, but what about the economic effects. What about the supressed interest rates and malivestment. How can we make it to the “mature” phase if every time the banks begin to shed reserves a business cycle happens?

    2.) Lets walk though an example. Suppose we are in the “mature” FR system. Bank A has 10,000 ounces of gold in reserve with liabilities of 100,000. A reserve ratio of 10%. It doesn’t expand any further because it believes it needs 10% reserves to protect against the risk of adverse clearing.

    Along comes a gold miner and deposits 1000 ounces of gold. The banks reserves now increase to 11,000 and its liabilites increase to 101,000. Yes, you are correct to say the banks liabilties have increased, but they have increased by a lower percentage than its assests. Thus the bank now has 11,000/101,000 = 10.89% reserves. Since it only needs 10% to avoid adverse clearing, the bank will expand credit by issuing 9000 unbacked notes. And of course these notes enter through the loan markets surpressing interest rates.

  9. Gene Callahan

    “What about the supressed interest rates and malivestment. How can we make it to the “mature” phase if every time the banks begin to shed reserves a business cycle happens?”

    What an absurd question! We ALREADY have FRB, so we don’t need to “make it” into the mature phase!

  10. Chris

    Gene, I’m speculating about the devolopmnt of banking under laissez faire. We’ve had state intervention in banking since the founding of this country. I would hardly consider our banking system as empirical evidence. If you want to bring up Scotland or Canada as examples and show that they developed under laissez faire and didn’t suffer from the problems I’m discussing. Fair enough, but those historical cases are another debate. And just pointing to those countries doesn’t say anything about why my theory is wrong, if it is wrong.

  11. Chris

    To elaborate a bit on point #4, If banks did pay interest on notes such that p(L)-interest > F I would find it to be an extraordinary coincidence if banks all paid exactly the same rate given differing investment opportunities available to each bank (and different portfolio risk). If the rate paid on notes is different from one bank to another then this significantly decreases FR notes viability as a medium of exchange.

    Furthermore, banks rarely pay interest on demand deposits. Why is this? Selgin writes of a diary entry from 17th century England which suggests that some early banks may have paid interest, but notes that this practice seems to have ended by the last two decades of the 17th century (how widespread this practice was is left up to our imaginations).

    The typical explanation is that the “free” account is your compensation for loaning your money to the bank. That seems plausible, but why doesn’t this compensation increase when rates increase? If the storage costs plus the maintenance on a 100% reserve account would cost me say 1%, and if interest rates were otherwise 1%, then we could say, yea the “free” account is my compensation for the loan to the bank. But if rates are say 5%, then competition should bid the rate on the demand deposit up to at least 4% (the 5% interest rate minus the value of the “free” account), correct? But we never see this happening. If rates go up to 20% all you get by the bank is a “free” account that you value at say 1%. The amount the bank pays never changes. Also, it doesn’t matter if I “loan” the bank $100,000 or $100 my compensation never changes… its just a free bank account. This seems awfully fishy to me. Also, how come when I put money in savings account I get interest but I don’t get any with a checking account despite that fact that these accounts are practically the same with only minor differences? One possible explanation is that people realize that when they put money in a savings account they are making a loan to the bank and so thy demand interest, but they don’t realize this with checking accounts so they don’t demand interest.

    I would tend to expect that if there were a “meeting of the minds” between the depositor and the bank that interest would be paid on FR accounts and those rates would fluctuate with interest rates in general. This would mean that FR notes would trade at different values relative to par again decreasing the possibly of use as a medium of exchange.

    All told I still remain very skeptical

  12. Jonathan Finegold Catalán Post author


    1. No, you misunderstand the mechanism. Let’s say we have 1,000 units of gold and 1,000 units of money-certificates, so 1 bank note = 1 unit of gold. Now, the bank sells an excess of 200 units of gold. The supply of fiduciary media is 200 notes, not 200 + 200, so the total supply of notes in circulation is still 1,000. Let’s say that instead of selling gold the bank decides to issue 200 notes, each still redeemable in 1 unit of gold. If this is unstable (i.e. it leads to a higher frequency/volume of adverse clearings) then this is a check on the system. In short, selling excess reserves of outside money is not inflationary; what is inflationary is excess note circulation, but we already knew this.

    2. Like I said in my original response to you, reserves do not dictate the extent of endogenous money creation! Your example presupposes that they do, and this is wrong. The 1:10 reserve ratio in your example assumes the original supply of money substitute and the original reserves of outside money. The ratio doesn’t operate independently from the supply of these different forms of money.

    4. You write,

    If the rate paid on notes is different from one bank to another then this significantly decreases FR notes viability as a medium of exchange.

    I don’t see why this is true. It just means that differences in p(L) – i will create competition between different fractional reserve banks. In a monopolistic environment banks don’t have to offer interest on demand deposits if fractional reserve banking has been institutionalized. This is also true if full reserve banking is no longer competitive; now fractional reserve banks only have to compete between each other.

    In our current system, banks offer interest on savings because they think it ties up your money for a longer period of time than a demand deposit. The idea is to decrease the volume of maturity mismatching.

    Regarding your last paragraph, the interest is paid on the deposit not on the note!

  13. Chris

    I must be dense because I’m not seeing it.

    1) “Let’s say we have 1,000 units of gold and 1,000 units of money-certificates, so 1 bank note = 1 unit of gold.”

    OK so far.

    “Now, the bank sells an excess of 200 units of gold. The supply of fiduciary media is 200 notes, not 200 + 200, so the total supply of notes in circulation is still 1,000.”

    First of all what do you mean to “sell” an excess unit of gold? Do you mean it loans out 200 of the 1000 units of gold it has in the vault? If this is the case you’re right the supply of notes is still 1,000, but now you’ve just placed 200 gold coins (units of gold as you called them) back into circulation. The total money in circulation is 1,000 notes + 200 coins = 1,200.

    Now what is going to happen from here is the borrower will take those 200 coins and buy something with them. The merchant on the other end of the sale will then take the 200 coins to his bank and deposit them in exchange for notes. Now the money supply is 1,200 notes.

    “Let’s say that instead of selling gold the bank decides to issue 200 notes, each still redeemable in 1 unit of gold. If this is unstable (i.e. it leads to a higher frequency/volume of adverse clearings) then this is a check on the system.”

    But this no different then when he lent out 200 units of gold in the first example. Both examples end up in the same place, with the money supply 200 units greater than it was before the bank lent out the money.

    I completely get the theory that in a mature FR system that banks wont expand credit because it will lead to a higher risk of adverse clearing. I’m not really questioning that. But it seems as if that analysis sort of assumes that a FR system just drops out of the sky in its mature form. Even if there is never any 100% reserve notes used at all, there still needs to be a transition from gold coins to FR notes/deposits backed by less than 100% reserve. I don’t see how this transition process wont be inflationary. If a bank believes it can hold 10% reserves without a risk of adverse clearing, and it is currently holding 100% reserves. It will loan out 90% of its reserves and start the process of inflation. It will only stop once the bank (and all others) reach the level of precautionary reserves they wish to hold. Its at this point it is said that the free banking system will act like a hard money system. But I’m arguing that it will never reach this point because the inflation during the transition process will generate the business cycle which will lead to a general deflation of the downside of the cycle.

    2) “Like I said in my original response to you, reserves do not dictate the extent of endogenous money creation! Your example presupposes that they do, and this is wrong. The 1:10 reserve ratio in your example assumes the original supply of money substitute and the original reserves of outside money. The ratio doesn’t operate independently from the supply of these different forms of money.”

    I don’t understand what you are saying here at all. The only thing constraining credit expansion is risk of adverse clearing correct? Unless I’ve been misreading Selgin this whole time I’m pretty sure that is the constraint.

    Now if a bank feels it is unprofitable to expand credit any further because of a risk of adverse clearing, then all of a sudden a gold miner deposits more reserves, the only way the bank would not expand credit in this situation is if the risk of adverse clearing has increased. Correct? Because the bank needs these new reserves to cover its increased risk of adverse clearing.

    But how can this be? What is driving the increase in adverse clearing risk? I suppose an increase in velocity would increase this risk, but this seems unlikely. Why would gold mining activity and upward changes in velocity be correlated?

    Given a constant velocity, I would argue the risk has actually gone down not up. The bank has more reserves as a percentage of liabilities, hence less risk of adverse clearing. It will find it profitable to expand credit in this situation. I can’t see it any other way.

    4.) “I don’t see why this is true. It just means that differences in p(L) – i will create competition between different fractional reserve banks.”

    I really doubt heterogeneous bank notes with different valuations will gain any traction as money. Banks would need to coordinate their efforts enough to produce homogeneous notes to gain widespread acceptance. In a competitive market this seems difficult to unlikely.

  14. Jonathan Finegold Catalán Post author


    (1) This is in the basic literature. It costs money to maintain gold in reserves. Excess gold is sold for non-monetary use. It’s not “put back into circulation,” precisely because people prefer to use bank notes. This is what you’ve been missing — the process of “shedding” excess reserves has nothing to do with loaning out money.

    (2) Right. The amount of outside money a bank holds in reserve is a function of the amount of outstanding liabilities. Just because someone deposits outside money and receives bank notes in exchange doesn’t meant the bank can pyramid on top of these reserves. What if two days later the note holder returns for the gold? You are misunderstanding the mechanism which restrains credit expansion, relying on this mystic notion of a “reserve ratio” which actually has nothing to do with anything. This isn’t just FRFB, but basic banking theory.

    (4) Again, interest is paid on deposits, not on notes. The value of a note is in how much gold it is redeemable for, so one “unit” is equal to one “unit.”

  15. Chris

    1.) Ok I get it finally. Wow! I have to admit I feel like I’ve read a lot of banking literature (and I have a Masters in finance!) and I’ve never once heard fractional reserve banking described this way. It is always taught that banks will loan excess reserves into circulation (gold standard or no gold standard) not simply sell them for non-monetary uses. So ya, as you describe it here it would not be inflationary. I’m not sure this is what anyone else has in mind when describing fractional reserve banking. And I just re-read the relevant chapters in The Theory of Free Banking i’m 99% positive Selgin is not describing FR banking the way you just did.

    2.) “Just because someone deposits outside money and receives bank notes in exchange doesn’t meant the bank can pyramid on top of these reserves.”

    I’m guessing you mean the bank would just sell this excess gold for non-monetary uses?!? This would be consistent with the way you just described FRB, but again, that’s not conventional FRB. And what is stopping them from just loaning the excess gold into circulation versus just “shedding” it?

    “You are misunderstanding the mechanism which restrains credit expansion relying on this mystic notion of a “reserve ratio””

    No I’m not! I have said specifically that the restraining mechanism that Selgin describes is the risk of adverse clearing. Which has gone DOWN if reserves go up relative to liabilities.

    1. Jonathan Finegold Catalán Post author

      1. Selgin (1988), pp. 21–22,

      Obviously it reduces the demand for coin in circulation, while generating a much smaller increase in the demand for coin in bank reserves. The net fall in demand creates a surplus of coin and bullion, which Ruritania may export or employ in some nonmonetary use. The result is an increased fulfillment of Ruritania’s nonmonetary desires with no sacrifice of its nonmonetary needs.

      Related to #2, continuing on p. 22,

      This causes a fall in the value of money, which in turn “acts as a brake” on the production of commodity money and directs factors of production to more urgent purposes (Wicksell 1935, 124).

      Back to the original point, p. 25,

      Even complete displacement of commodity money in circulation by inside money does not, however, necessarily mean increased fiduciary substitution. Commodity money, formerly used in circulation to settle exchanges outside Ruritania’s banks, might now be used to settle clearings among them… Then only net clearings, rather than gross clearings, need to be settled in commodity money. Thus banks can take further advantage of the law of large numbers, and more commodity money becomes available for nonmonetary uses.

      How I describe this process is actually taken from Selgin (1988), so it would surprise me if Selgin describes the process substantially differently.

      2. What is “conventional FRB?” We’re talking about a free banking system (and, not even a cartelized FRB system works by “pyramiding on reserves” — this is something that a lot of Austrians, probably influenced by Rothbard, repeat a lot, but it doesn’t actually describe endogenous money creation; not to mention that banks need to hold reserves to meet other types of liabilities). And no, I didn’t mean that the bank would sell excess gold. What I meant is that a bank would only be able to increase the supply of fiduciary media, based on the new deposit of outside money, if the depositor’s demand for money rises. The reserve ratio isn’t independent of the supply of outside or inside money; it’s based on on these factors and the behavior of note holders.

      I will repeat, the “risk of adverse clearings” does not go down just because reserves increased. That marginal amount of reserves (the new deposit) is subject to the demand for its redemption by the holder of the money-certificates.

  16. Chris

    I think you’ve really taken the liberty of interpreting Selgin in a way he wasn’t intending. Two pages earlier Selgin writes:

    “Of course, Ruritania’s primitive bankers may also engage in lending;
    but their loans are originally made out of their personal wealth
    and revenues. The lending of depositors’ balances is a significant
    innovation: it taps a vast new source of loanable funds and fundamentally
    alters the relationship between Ruritanian bankers and their
    depositors. “The … bailee develops into the debtor of the depositor;
    and the depositor becomes an investor who loans his money … for a

    There’s no talk of “shedding” reserves here. Plus, in the paragraph you quote he is just suggesting that the fall in demand for gold in exchange may allow some additional gold to flow to non-monetary uses. Which is fairly obvious. But you’ve taken a giant leap to suggest that means banks will move from 100% reserves to fractional reserves only by “shedding” for non-monetary purposes leaving the money supply unchanged and never by loaning money into circulation. I would tend to think the latter process would vastly outweigh the former, by the process described by selgin in the paragraph I quoted.

    2.) I still think you’re missing something. Lets go back to the mechanism that restrains credit expansion. How does the clearing mechanism work? Bank A issues notes and deposits to its customers. Its customers buy things from customers of Bank B. The purchases are made with notes issued by Bank A, checks, and debit transactions. Bank B piles up claims on Bank A which it submits to the clearinghouse. The clearinghouse then charges Bank A for the claims of Bank B (and the claims of other banks as well).

    But at the same time Bank A is also racking up claims on Bank B (and other banks) due the customers of the other banks buying things from Bank A customers and paying with notes, checks, debits etc. In theory the amount of claims Bank A owes to other banks would be offset by the amount of claims owed to it by other banks. On the net, Bank A breaks even.

    However, in any given clearing period there is no guarantee that bank A will net out even at the clearinghouse. There is a very real possibility that it could actually have a negative balance at the end of the clearing period, if for one reason or another, competitor banks rack up a greater amount of claims against it then it has against them. In the long run this should work out to break even but in any given period it may have a negative or positive balance. (It should be noted that there is a serious incentive here for banks to cooperate to lengthen the clearing period to engender credit expansion.) In order to guard against having a negative clearing balance (an adverse clearing event) the bank holds precautionary reserves. The bank uses its financial know-how to speculate as to how much reserves it needs to avoid an adverse clearing event and strives to maintain this target. If it has excess reserves, it loans them out. If it is under on reserves it will contract its outstanding liabilities to maintain this target.

    Now we can see that the level of reserves is contingent on the probability of having a negative balance in any given clearing period. Increases in the demand for money reduce the volume of clearing transactions, reducing the probability of a negative clearing balance. While decreases in the demand for money increase the volume of clearing transactions increasing the risk of a negative clearing balance. The FR banks, presumably, will expand and contract money and credit as the risk of an adverse clearing event changes due to the change in the volume of clearing transactions which is ultimately dependent on the demand for money.

    But if the demand for money remains unchanged. The volume of clearing transactions also remains unchanged. And it follows the chance of incurring a negative clearing balance also remains unchanged. Thus the bank will leave its reserve target unchanged. Given this framework what happens if a gold miner deposits newly mined gold at the bank? Technically, since the bank will issue notes or deposits to the miner, the volume of clearing transactions will increase slightly and the risk of a negative clearing balance will increase slightly. Thus the bank needs to carry a slightly greater amount of reserves. But its reserves have increased in the form of the newly mined gold. And not only have they increased enough to cover this increased risk, but the bank now has excess reserves (since its reserves increased by a greater percentage than its liabilities). And it will lend out those excess reserves. There is nothing stopping it from doing so. It doesn’t need them to cover any risk of a negative clearing balance. Credit expansion, not backed by an increase in monetary demand will take place.

    I read an article recently by either Selgin and White which suggested that banks have historically operated with reserves as low as 3%. If we assume a .5% increase in the gold stock mined each year, and 3% reserves, that translates into about a 17% increase in the money supply each year. All filtering through the credit markets pushing down interest rates. There is no way this would be sustainable.

    1. Jonathan Finegold Catalán Post author

      1. That quote is discussing something completely different! Banks do lend from deposits, but this lending is regulated by the demand for money and the risk of adverse clearing. This is separate from the idea of excess reserves of outside money. But, as it turns out, you have it backwards — the major force is a reduction in reserves, not a rise in fiduciary media (largely because the demand for money should be relatively stable). Read the entire chapter by Selgin — there’s a lot of things being discussed in it (so it makes sense that he’s talking about something different somewhere else in that chapter!).

      2. I don’t know how our discussion got to inter-bank clearings. I guess it might have something to do with that second quote I posted. That wasn’t what I was drawing your attention to, which is why I skipped over much of it — it was just an aside on the forces which would limit gold production for monetary purposes (since that’s what we’re talking about).

      In any case, the demand for money doesn’t remain unchanged. The new depositor’s demand for outside money doesn’t suddenly become tied to the social demand for outside money!

      And you keep repeating the argument about how the supply of fiduciary media is dictated by the “reserve ratio.” This is nonsense! The relation is the exact opposite! The reserve ratio is dictated by the supply of fiduciary media!

  17. Chris

    You keep telling me I’m wrong g without actually providing it!

    I went into detail to describe how I interpret the forces Selgin is describing. The only force serving as a constraint on fiduciary media creation is the possibility of a negative balance in the clearing mechanism. That’s it. The degree to which the demand for money effects fiduciary creation is manifested through this clearing mechanism, not independently of it.

    I’ve been told I’m wrong, but you have provided any mechanism showing where or where I’m wrong. Just saying it doesn’t prove it.

    1. Jonathan Finegold Catalán Post author

      Err, right, but the “possibility of a negative balance in the clearing mechanism” changes with changes in preference of old and new depositors. I’m not sure what’s so hard to understand about this.

  18. Chris

    Let me give a final crack at this. In Selgin’s model, the only thing prentiving total unlimited credit expansion is the possibility of running a negative clearing balance during any given clearing period. It is to guard against this possibility that banks self-impose a reserve requirement. If the possibility for a negative clearing balance wasn’t there, we would see unlimited credit expansion. I focused on the clearinghouse because this is THE key to Selgin’s system. If the risk of a negative clearing balance decreases, banks expand credit. If it increases, banks contract credit.

    I believe it was Bagus and Howden in one of their papers where they suggested that if the banks cooperated and extended the clearing period, this would engender credit expansion because over a longer clearing period the probability of a negative clearing balance shrinks. This is because inflows and outflows should net out to zero in the long run. Technically, if banks cooperated in extending the clearing period indefinately, then the check on credit expansion would be removed entirely and there would be unlimited inflation. George Selgin in his reponse to them basically said, ya you’re right, but I don’t think this is likely. But nonetheless, this is key to understand this, if the reserves a banks needs to hold falls, the bank will lend out any excess reserves. Selgin clearly understands this. There are probably many places in his writings where this is evident, but that one comes to mind.

    Lets give an example. Suppose the demand for money increases. This increased demand manifests itself through the CLEARING mechanism. The result is the volume of clearing transactions decreases and the risk of having a negative balance in any given period decreases. Thus, the bank will lower its self imposed reserve requirement. At this point the bank is now holding EXCESS RESERVES! The bank proceeds to loan out these excess reserves expanding credit (which coresponds to an increase in the demand for money). The increased demand for money is not technically what caused the credit expansion. It was the the excess reserves (brought about by the increased demand for money) that caused the credit expansion.

    Thus, if reserves increase due to gold mining. The banks will loan them out in the exact same fashion. The profitability of holding money doesn’t have anything to do with whether banks expand credit or not. Everything hinges on the level of reserves needed to protect against adverse clearing.

    1. Jonathan Finegold Catalán Post author

      We’ll have to agree to disagree, but a final point: you are clearly confusing demand for inside with demand for outside money, and you are clearly confusing reserves (i.e. idle deposits) of inside money with reserves of outside money. Yes, a rise in demand for money will increase the supply of idle inside money; this is not the same mechanism which influences the use of outside money.

  19. Gene Callahan

    Well, Chris, what does starting from a mythical state of laissez faire have to do with a recommendation to move to free banking from where we really are?

  20. Pingback: Some Words on 1893 Australia | Economic Thought

Leave a Reply