This archived content originally appeared at, the predecessor site to, and does not carry the sponsorship of the Cato Institute.

Have we got your back?

Here is my long promised, though not likely long anticipated, post on the “backing” theory of money.

The modern controversy over "backing" originates from this article by Neil Wallace and Thomas Sargent. In the article they offer “ a simple model that is compatible with the principle of finance theory that assets are valued according to the streams of returns that back them” and appear to endorse that principle as a basis for explaining the value of money.

“Backing” clearly applies to certain cases but to my mind it just as clearly cannot serve as a general explanation of the value of the monetary base without stretching the idea so far as to make it a metaphor so vague as to have little use.

Currency boards are cases where the idea of backing is useful. Currency boards are established to provide credibility for the local currency by backing their notes, coins, and deposit liabilities (if any) 100 percent or slightly more in high-quality foreign assets. In principle, 100 percent foreign reserve backing means that a currency board can liquidate almost immediately and exchange all of its notes, coins and deposits for foreign currency at the fixed exchange rate it maintains.

For central banks, the idea of backing is less useful as an explanation of why the currencies they issue have value. Central banks can become technically insolvent with little effect on the currency, as some in fact have. (For a discussion of central bank insolvency, see this paper by Willem Buiter. Maxwell Fry’s good book Money, Interest and Banking in Economic Development also treats the subject.) As an economist I met years ago said of such cases, “The assets of the central bank are garbage; the liabilities, everyone believes in.”

There are also cases where fiat monies continue to circulate and retain value even in the absence of any politically powerful issue. Such cases happen most often during or just after wartime, when, for instance, an occupying army that has lost the war withdraws and leaves its currency behind. For example, after World War I war ended, German occupation currency continued to circulate in what are now Poland and the Baltic states during a kind of interregnum. In such cases, if the new government does nothing to hinder the currency’s use, the currency can continue to circulate even though the original issuer will not redeem it for anything.

One can reply that in the cases of insolvent central banks, occupation currencies whose occupiers have left the scene, or other examples that may be given, the currency retains value because it is either backed by the assets of the whole of the government, not just the central bank, or that in the case of an occupation currency that people expect it to receive such backing from the new ruling government. To me, that is stretching the metaphor of backing to the point of useless vagueness. A fiat currency accepted by the government in payment of taxes is not “backed” by tax revenues or even by the general assets of the government in the sense that a currency board is backed by its reserves. Holders of the fiat currency cannot necessarily redeem it for any external asset at a set rate upon liquidation.

What is the alternative, then? It is that value is based on acceptability. An unbacked currency can retain its value if people continue, for whatever reason, to accept it in payment and to hold onto it as a store of value. Conversely, a fully backed currency may not be acceptable and may not serve as a true money. As I have mentioned in previous posts, it is legal for U.S. banks to issue notes, but I cannot imagine that a bank note issue denominated in Argentine pesos, fully backed with Argentine government bonds or even with Swiss bonds segregated from the bank’s other assets, would circulate in the United States.

* * * * * * * * *

Mike Sproul has the right of a guest post in reply if he wishes to use it. After that, I may or may not offer a final comment. I am now busy with a book (on currency boards, not free banking) and other projects and will be for many months, so instead of posting approximately weekly, I will be far less frequent. Without intending it I became the most frequent contributor to the blog. The editors have received my ideas for continuing a flow of content, but it is up to them whether to keep up the flow or let the blog peter out.

  • MichaelM

    I sure hope this blog doesn't peter out. I suppose it says something to the niche interest this specific topic generates that there is a worry of that. The only solace I can see there is a comparison between the minor niche the topic occupies today and the miniscule existence it had back in the 80's, when (to my understanding) the modern free banking school was little more than a half dozen economists and economics students with wild ideas about what was and what could be.

    The several elephants in the room that seem to generate the most interest (discussions on bitcoins/cryptocurrency, discussions on 100% gold reserve banking, etc), unfortunately, seem to be the most limelight the MFBS is going to get for now. I just hope the future continues to be just a little bit brighter for this most interesting of econo-policy thinker-groups.

  • Paul Marks

    I do not like the language of "backing" – it confuses the issue (and thus opens the door to deception – to fraud, as an ordinary person would understand "fraud").

    Either the commodity (whether it is gold, silver, copper – whatever the commodity is) should be the money – or it should be clearly stated that it is NOT the money.

    A situation such as the 1920s where it was pretended that gold was the money, but really a vast credit bubble was being actively pushed (by Benjamin Strong of the New York Federal Reserve – and his friend M. Norman Governor of the Bank of England).

    When people in India asked for real gold-as-money this was actively sabotaged by the Bank of England – and both the British and American authorities backed credit bubbleism all over Latin America – indeed as far away as China (where the traditional silver-as-money was systematically, and deliberately, undermined).

    Paul Johnson (in "Modern Times") was right to call the monetary system of the 1920s not a "gold standard" but a "not in front of the children standard" – a deception, and a deception on a truly incredible scale.

  • It might be useful to think of backing as one type of proof of scarcity. A sine qua non for acceptability is that the acceptor of a currency believe he will not be met in the marketplace with barrels of the stuff bidding against him. Any of the following beliefs will provide the requisite confidence in scarcity:

    That a coin contains something valuable
    That paper was issued only in exchange for something of value
    That paper can be exchanged for something of value
    That paper was issued under a political regime in which the over-issuance of money costs people their jobs
    That paper was issued in an information-age environment where the failure of the inflation hawk to squawk is evidence of scarcity.

    Perhaps other bases for believing in scarcity are available, too. The great leap forward represented by fiat currency is that it allows scarcity to be achieved dynamically, by reference to the available supply of goods and services, rather than by reference to the acquisition of special dirt. In a well-run fiat regime, there will be a satisfactory correlation, inherent in the way money is created, e.g., by banks assessing creditworthiness, between the amount of money created and the amount value of the outputs that will be offered by reason of its creation. (In other words, if I can persuade a bank to lend me the capital to manufacture something, there is a pretty good bet that I will produce enough stuff to absorb that money when it leaves my hands and starts looking for goods.)

    In this context, backing becomes just another technology, one especially suited to a certain level of information technology somewhere between the printing press and the television. But there is nothing intrinsically special about it vis a vis any other technology for establishing that a currency is scarce enough to be money.

  • Mike Sproul

    We agree that the backing theory gives a useful description of the operation of currency boards, so I'll focus on that for now. In its idealized form, a currency board might have issued 100 paper pesos in exchange for $100 of federal reserve notes (FRN's). The board would maintain convertibility at $1=1 peso. If the board issued another 200 pesos in exchange for another $200, then there would be $300 of FRN's backing 300 pesos, and 1 peso would still be worth $1. A quantity theorist might have expected the tripling of the quantity of pesos to reduce the value of the peso by 2/3, but this is clearly a mis-application of the quantity theory. The value of the peso must stay at $1, just as the backing theory implies.

    Now let's allow our currency board to sell $200 of FRN's for $200 of US government bonds, so as to earn some interest. Is it still a currency board, or should we call it a bank? Kurt seems to say above that it still can be called a currency board. Does the backing theory still hold? Yes, it does. Whether the board holds $100 worth of FRN's and bonds as backing for 100 pesos, or $300 worth of FRN's and bonds as backing for 300 pesos, 1 peso is still worth $1. The backing theory is right again, and the quantity theory is wrong again. (Plus the bank–ahem–currency board, now has a source of interest income to pay its bills.)

    What if our currency board suspended convertibility of pesos into FRN's? Surely we must now call it a central bank. But wait. Every currency board suspends FRN convertibility at night and on weekends, so that kind of suspension doesn't matter. What if the currency board promised FRN convertibility once a month, or once a year, or once a century? Isn't that still a kind of convertibility? Of course it is. The rate of interest will play a bigger role as the time to conversion increases, but the fundamental truth remains: If the currency board triples the amount of pesos, while at the same time its assets triple, then it remains true that 1 peso=$1, and once again the backing theory is right while the quantity theory is wrong.

    CONCLUSION: Currency boards and central banks are just slight variations of each other. Once you see that the backing theory correctly describes the operations of a currency board, it becomes clear that the same is true of a central bank.

    • Kurt Schuler

      Here's a kicker: At least three currency boards have had negative net worth for long periods. Argentina's Caja de Conversion of a century ago was established to allow a fiduciary issue of 293 million pesos; only issues above that amount required 100% reserves in gold. The East African Currency Board had substantial negative net worth from its start in 1921 through 1944 because the British colonial authorities made a bad exchange rate mistake with old currency at the start (involving the value of silver coins). Hong Kong's Exchange Fund had modestly negative net worth from 1946-1948 as the result of a government decision to honor notes issued fraudulently by the Japanese occupation authorities during World War II. The only one of these cases in which any currency depreciation occurred was in Argentina, at the start of World War I, when every other major country was officially abandoning the gold standard or imposing such obstacles that foreign exchange was in effect controlled.

      Why didn't Argentina's currency depreciate against its gold anchor before 1914, and why didn't the East African shilling and the Hong Kong dollar never depreciate against their anchor, the pound sterling? It was well understood by administrators of currency boards and economists who studied them that there is a "hard core" of circulation that is highly unlikely to be redeemed; it takes something shocking and unexpected to reduce it, such as a war engulfing the whole of Europe. The hard core, I submit, is the result of the wide acceptability of the currency. So I agree that there are many similarities between central banks and currency boards, but both my antecedent thoughts and the conclusions I draw differ from yours.

      • Mike Sproul

        The backing theory explanation for the survival of insolvent currency boards is that the public must have believed that the currency board would ultimately be bailed out, either by the government, by the IMF, or by the US. History shows this belief to be well-founded.

        If a currency board were truly insolvent, then it would be an arbitrage opportunity waiting to be plucked by speculators. If, for example, the currency board holds $90 of assets against 100 pesos that it has issued, and if there is no Santa Claus standing by, then 1 peso=$.90. If the currency board tries to maintain convertibility at 1 peso=$1, then net worth is negative. The first 90 pesos to be redeemed will get $1 each, and the last 10 pesos will get nothing, so a run is inevitable, hard core or no hard core. Also, short sellers will get in on the act, borrowing pesos and selling them for dollars, and putting themselves in a position to profit as the peso falls. Meanwhile, what is so "hard core" about the hard core? People always start using substitute moneys, foreign currency, etc. when there own currency fails.

        • Kurt Schuler

          One might argue that for the Hong Kong Exchange Fund negative net worth was small and for the East African Currency Board, the governments involved did pledge to back it after about a decade when its foreign reserves fell to less than 10% of notes and currency in circulation. (That turned out to be the low point of the Great Depression and the low point of reserves.) The Argentine Caja de Conversion was definitely not backed by the other resources of the government and given Argentina's history the public did not seem to have had any expectation that it would be. So at least for it the bailout hypothesis does not hold.

          • Mike Sproul

            Assume you're right about Argentina: absolutely zero chance of anyone coming to the rescue. If they maintain convertibility at 1 peso=$1, when their resources are only enough to support 1 peso=$.90, how can such a glaring arbitrage opportunity survive a speculative attack?

            (I'm willing to bet that there was some hidden complication to the Argentine case that somehow did provide backing. When the facts seem to fly in the face of the no-arbitrage principle, I tend to suspect the "facts" and go with principle.)

          • Kurt Schuler

            (Replying here to Mike's comment below because the comment software apparently will not allow any further sublevels of replies.) I have read about this episode extensively. The reason there wasn't a run on the Caja de Conversion from the time it started accumulating gold in lasting fashion (January 1903) until World War I was that people wanted paper money to make certain kinds of payments, and the Caja de Conversion had a monopoly of issue. That created a hard core of demand for its notes. Over time, growth in the Argentine economy made the hard core larger and larger. There was a 100% reserve backing requirement for pesos issued beyond the fiduciary issue of 293 million pesos. Gold reserves rose from zero in early 1900 to more than 60% before the war.

          • Mike Sproul

            How did the Caja de Conversion put money into circulation? Surely it did not simply give away the first few million paper notes (pesos?), and only later start getting gold in return for them. So let me make an educated guess: The first few notes were issued to pay the salaries of government workers, to buy goods for government offices, etc. This is exactly how American colonial paper money was issued, and in every case, the notes were acceptable in payment of taxes. I'd guess that this must have been the case for the pesos issued for the Caja de Conversion. Otherwise, why and how were the pesos issued?

            Also: Quantity theorists often claim that the Massachusetts shillings of 1690 were unbacked, and they point to the colony's empty treasury in support of this claim. They fail to see that "taxes receivable" is a real asset to the government, and can serve as backing to the money issued by that government.

          • Kurt Schuler

            Yes, the pesos were issued to pay government expenses. Later the Argentine government wished to make the peso a gold-standard currency, so it froze the supply of peso notes and imposed a 100% gold reserve requirement for new notes: if you wanted to get pesos beyond the frozen amount, you had to pay gold to the government to issue them. Your comment goes to the heart of the issue for me, which is that I consider the term "backing" to be misleading for a floating currency. There is no external asset that is voluntarily exchanged–nothing to get back. The term "acceptance" is more accurate, and "forced acceptance" still more accurate. In such cases, the government accepts for payment of taxes the currency it has previously issued. It's a forced exchange: people who don't pay the taxes go to jail; depending on the law, they may also go to jail for refusing to accept the currency in payments not involving the government. So, to count the forced acceptance of a floating currency in taxes as a kind of asset is rather like me putting my fists down on my personal balance sheet as an asset because I could use them to punch a cyclist and ride away on his bike. But we're going on too long in the comments here when I think it would be more useful for us to write real posts, which I invite you to do when you have time.

        • Mike Sproul


          I'll take your advice and submit a complete post, probably in 2-3 weeks. But let me propose a thought experiment first: A landlord collects rent of 50 oz of silver per year in perpetuity. At an interest rate of 5%, this makes his land worth 1000 oz. So we could say his assets consist of land worth 1000 oz, or alternatively, his assets consist of 1000 oz worth of "rents receivable". if he has no debts, his net worth is 1000 oz.

          The landlord could pay for his groceries by handing the grocer a slip of paper that says "OK for 1 oz rent on my property". The grocer might accept the slip if he rents from the landlord or does business with someone who does. The landlord could issue up to 1000 of those slips and spend them, and they might circulate as money, adequately backed by the 1000 oz of rents receivable. The landlord might have gotten his land by theft, and he might threaten to imprison tenants who don't pay, but that's beside the point. His slips are backed by "rents receivable", regardless of whether he's a nice guy.

          Change "landlord" to "Argentine government", and change "rents receivable" to "taxes receivable", and the Argentine episode is explained.

          At this point the landlord could freeze the quantity of slips and declare that new slips will only be issued to those who deposit 1 oz of silver. No problem. As each new slip is issued, he gets another oz, so 1 slip is still worth 1 oz. Plus he now has some silver, so he could start to offer "silver convertibility" in addition to the "rent convertibility" that he already offers. If those silver-issued slips are all redeemed for silver, so that he's back to having 1000 oz of rents backing 1000 slips, then 1 slip is still worth 1 oz, but quantity theorists might make the mistake of thinking "inconvertible"="unbacked", and declare that the slips are now unbacked "fiat" money.

  • ivancarrino

    Very interesting post Kurt! I have some questions:

    "a fully backed currency may not be acceptable and may not serve as a true money."

    How? Why would you refuse to accept a liability while accepting the asset with which it is backed?

    "As I have mentioned in previous posts, it is legal for U.S. banks to issue notes, but I cannot imagine that a bank note issue denominated in Argentine pesos, fully backed with Argentine government bonds"

    Great, but isn't this the case of a "bad backing"? I mean, of course I would refuse to accept Argentine bonds, that's a government in which nobody trusts….

  • Nicely done. Acceptability as a determinant of value is the same as saying value is subjective. As you point out, the issuers may exchange a note for a specified amount of foreign exchange or commodities, but people only care about that if they are going to hoard the money. If they intend to use the notes to buy goods, the exchange ratio of the notes per goods is more important.

    The official exchange rate per foreign exchange or metals is very much like the coupon interest rate for a bond. But the exchange value of the bond per other money changes radically depending on interest rates, expectations, and many other things. In the same way the exchange value of notes per goods will change radically even if the "coupon" rate of exchange set by the bank does not. Mike wants to talk only about the coupon rate of notes and ignore the exchange for goods value.

    • Mike Sproul

      The trouble with "Acceptability as a determinant of value" is rival moneys. If, for example, people start using some foreign money, then the demand for the domestic money falls, which reduces its value. The supposed free lunch that had been earned by the issuer of the domestic money is transferred to the issuer of the foreign money. Given this rivalry, foreign moneys would continue to invade the domestic circulation until the value of the domestic currency fell to zero.

      The backing theory does not have that problem, as the issuer earns no free lunch, and there is no profit to the issuers of rival moneys.

  • George,

    I suggest your concluding (2nd last) paragraph is 75% right rather than 100% right. You say “What is the alternative, then? It is that value is based on acceptability. An unbacked currency can retain its value if people continue, for whatever reason, to accept it in payment and to hold onto it as a store of value.”

    I don’t like the phrase “for whatever reason”. The “reason” people want a currency is very simple. It’s the reason set out in the introductory economics text books, namely that money is useful: it obviates the relatively inefficient alternative, namely barter. Everything else is irrelevant by comparison.

    E.g. a currency does not necessarily have to be issued by a bank (central or commercial). It is not necessary for government to demand the currency in payment of taxes (though that doubtless helps). It does not need to be backed by any asset, like gold.

    I.e. it all really just boils down to the simple fact that money is useful. So people will accept and use ANY FORM of money, even relatively chaotic Mugabe types of money.

    • Ralph,

      Well said. And the common medium of exchange function isn't merely valuable to people, it is extremely valuable to almost all people who are engaged in a marketplace. And the extreme value is understandable in light of the phenomenally large increase in trade efficiency that a money permits.

      And I don't believe my choice of words here are in the least bit hyperbolic. As such, the real peculiarity, even in light of the obvious hurdle from obscurity to widespread acceptance, would be if a money did NOT emerge in a large society.