Showing posts with label Regression theorem. Show all posts
Showing posts with label Regression theorem. Show all posts

Saturday, April 6, 2013

Bitcoin is no Great Mystery

To cut a long story short, Bitcoin (or BTC) looks more and more like a quasi-financial asset in a bubble every day to me.

And Bitcoin is nothing like a real asset like a house or property; and it only roughly fulfils the definition of a financial asset. But a quasi-stock market is a good analogy for one main activity associated with Bitcoins: the exchange of them for local currencies.

Some intellectually interesting points are these: is a Bitcoin some kind of “produced” good, or more like a type of financial asset; and does the Bitcoin as a means of payment violate Mises’s regression theorem?

Some quick background. Bitcoin is a “decentralized digital currency” arising from, and depending on, an open-source, peer-to-peer internet protocol. It made its appearance in 2009. Bitcoin has an ambitious claim to divisibility: one Bitcoin can be subdivided into 100 million smaller units called “satoshis.” The stock of Bitcoins is increased by people downloading certain software and (generally) joining Bitcoin “pools,” in which they earn Bitcoins by making their computers solve algorithms. But the new number of Bitcoins created in the future will diminish and drop off. Eventually, there will be a fixed and finite supply of 21 million Bitcoins. But, notably, there are already issues of fraud in the manipulation of Bitcoin software.

But, for those who are not tech savvy, creating Bitcoins looks like a confusing, time-consuming, and exhausting business, not worth the effort. (Also, I notice that to ensure that Bitcoins are not lost if your computer crashes, you must print out or have some copy of the “wallet.dat” file.)

There are about 11 million Bitcoins in circulation, according to this graph.

But, curiously, it is not the daily number of transactions in Bitcoins (which rose from about 4000 in late 2012 to 7000 in April 2013) that has soared recently, but the price of Bitcoins measured in other currencies.

Has price stability of goods available in Bitcoins been observed since 2009? Apparently not. Bitcoin land had a kind of hyperinflation in 2011, set off by the bursting of the first bubble and loss of confidence when hackers who stole many Bitcoins. Now essentially deflation in goods priced in Bitcoins is going on. While supporters might argue the latter is a good thing for Bitcoin holders (sure, it is), that was not the issue at hand: did price stability result? No.

If Bitcoin ever wants to be a real currency in the future, it must consider the needs of hypothetical Bitcoin debtors. For example, would you want to take out a loan in Bitcoins and then get hit by debt deflation? And apparently real world Bitcoin debtors do exist (as to the scale of Bitcoin lending, I have no good data.) In short, with no relative price stability, there will be no significant demand for Bitcoin loans, so you can say goodbye to serious and economically viable debt/credit developments within Bitcoin!

There is also another issue about the nature of Bicoins. Is a Bitcoin a consumption good or was it originally a consumption good? No, it was not. People do not produce Bitcoins to just “consume” them on their computer. Certainly, Bitcoins are not like some perishable consumer good, in the way a carton of milk or loaf of bread is.

Are they a durable consumer good? That looks false too. What use is a Bitcoin if you just hold it, never intend to sell it or use it in some exchange? How does it function as durable consumer good? It does not. Nor is a Bitcoin a capital good.

There is a twofold explanation of what Bitcoins actually are. First, if the people who first held them expected to derive indirect utility from Bitcoins by buying goods with them in the future, then they were being treated as if they were money from their inception. (If people derived direct utility from them by thinking that Bitcoins would provide them with future liquidity, then they also fulfilled the function of money from their inception, though I think this is less likely.) Secondly, if people had the expectation that Bitcoins would rise in monetary value, then Bitcoins also look like a type of speculative asset, or, better, a speculative digital asset (as opposed to a real or financial asset), but one not tied to any company in the way a private stock or bond is.

Bitcoins were obviously created to be “decentralized digital currency,” a type of money to be used as a medium of exchange and store of value. That is how the Bitcoin software was “advertised,” if you like. But a Bitcoin is not tied to some real commodity like gold at a fixed conversion rate. One wonders why any libertarian would get excited about it (as it turns out, most do not!).

A Bitcoin is backed by no commodity whatsoever: just like a stock or share, whose value is subjective and whose price is just determined by supply and demand on a stock market. The value of Bitcoins in goods or other currencies might crash tomorrow (and so might gold, but at least we have industrial uses for gold and uses in jewellery, etc.).

At the moment, the supply of Bitcoins might be regarded as somewhat elastic, if subject to diminishing future elasticity, and eventually zero elasticity. So, in the long run, there will be a finite supply of Bitcoins and an inelastic supply.

Here is the crucial point: a money facing a long-run inelastic and fixed supply is not much use for a real world economy, in my opinion. It will be useful mostly as some kind of a speculative asset.

Perhaps some kind of Bitcoin fractional reserve banks might emerge offering what standard fractional reserve banks offer: a mutuum contract where you sell your Bitcoins to the bank in return for a financial instrument or Bitcoin “demand deposit” (that is, Bitcoin debt money). But any run on these banks would likely cause utter collapse, as the base money of Bitcoins is so inelastic. Who will be the lender of last resort in a Bitcoin bank run? The answer: nobody.

If one wants to be ultra-cynical, Bitcoins are just another speculative activity for those producing and buying them. It is just another type of gold buggery – except even more ridiculous. It also has the potential to be exploited by criminals, and is therefore of questionable social value.

The whole Bitcoin phenomenon is like a decentralised operation to produce quasi-financial assets like stocks or shares, except you get no ownership stake in anything (except individual Bitcoins) and no dividends, only an alleged digital “currency” of marginal use in purchasing real goods, whose value is subject to wild fluctuations just like stocks and shares. There might be some transactions demand and (more likely) a speculative demand for holding them, but so what?

If you want to say that Bitcoins are nothing but a quasi-financial asset (without even many of the other properties of a financial asset), then (from the libertarian perspective) Bitcoins are just inherently worthless digital code, just as the physical paper record of a promissory note or share is virtually worthless. But just like a share, you can expect value fluctuations in the price of a Bitcoin.

Now a different issue. Austrians in general seem quite hostile to Bitcoins. Why? The reason is that they appear to violate certain Austrian theories about money.

First, Menger’s theory of money’s origins requires that a commodity emerges in the market as a real and actively exchanged good in spot barter trades. It then attains a “virtually unlimited saleableness” and thereby emerges as the dominant medium of exchange. But Menger’s theory does not really say much about the emergence of new “money things” in a society already dominated by money.

While Menger’s theory of money may be of questionable relevance for the subject of Bitcoins (I am undecided on this question), Mises’s regression theorem looks relevant, especially the strident, dogmatic statements made by Austrians like Rothbard:
“[sc. Mises’s] Regression Theorem also shows that money, in any society, can only become established by a market process emerging from barter. Money cannot be established by a social contract, by government imposition, or by artificial schemes proposed by economists.” (Rothbard 2009: 61).
What does Rothbard mean by “money”? If one wants to take “money” as just a widely-used medium of exchange, then the existence of numerous transactions in which Bitcoins are exchanged directly for goods means that Bitcoins are becoming a type of money (and there seem to be many places where you can buy goods directly with Bitcoins).

But a barter process presupposes that any “money thing” was previously exchanged in spot trades as a good, whether a capital good or consumption good, before becoming money. Yet I find it impossible to accept that Bitcoins were an “economic good” in either sense before rising to the point of being used in exchange.

While I would not press my conclusions here too far (since I have not studied the Bitcoin transactions for real goods in great detail), it looks like Austrians should explain why the use of Bitcoins as money does not violate their Regression Theorem. (This is somewhat of a waste of time, however, given that the Regression Theorem is trying to solve a pseudo-problem. Money has direct utility, and its value does not consist just in its current purchasing power.)

In short, let me conclude by saying that heterodox Keynesians or advocates of MMT face no theoretical difficulty explaining the emergence of Bitcoins. Our theory says that money can be created by anyone (think of debt money such as negotiable bills of exchange, or promissory notes), but the problem is getting it accepted (Papadimitriou and Wray 2010: 9). Bitcoins, for various reasons, have won some degree of acceptance. But, without relative price stability and an elastic supply, Bitcoins are not a viable monetary unit for any large capitalist system.

Any serious form of money ultimately develops into mere “money base” in a fractional reserve banking system with further credit money pyramided on top of it (just as gold did). But a Bitcoin fractional reserve banking system is not viable without an elastic supply, nor would a system of lending denominated in Bitcoins be realistic without relative price stability.

And last but not least: if you cannot pay your taxes in Bitcoins, then Bitcoins will never, ever displace real-world, national fiat currencies.

Meanwhile you might like to gamble in Bitcoins, or for that matter in stocks or shares, or in gold, but you might lose your shirt too.


“Money Has Direct Utility,” October 25, 2012.

“Mises’s Regression Theorem: A Critique,” January 13, 2012.

Matias Vernengo, “More deflation in Bitcoinland,” Naked Keynesianism, February 19, 2013.

Matias Vernengo, “Hyperinflation in Bitcoinland,” Naked Keynesianism, January 1, 2012.


Papadimitriou, Dimitri B. and L. Randall Wray. 2010. “Introduction: Minsky on Money, Banking and Finance,”
in Dimitri B. Papadimitriou and L. Randall Wray (eds.), The Elgar Companion to Hyman Minsky. Edward Elgar, Cheltenham. 1-30.

Rothbard, M. N. 2009. The Essential von Mises. Ludwig von Mises Institute, Auburn, Alabama.

Thursday, October 25, 2012

Money Has Direct Utility

The idea that money only has utility through its exchange value – or, that is to say, money’s indirect utility is derived from the utility of any commodity or set of possible commodities you can purchase with it – is held by Austrians and, as far as I can see, neoclassicals.

This view is wrong: for money can yield direct utility (Graziani 2003: 11).

Money or even highly liquid financial assets (like deposit-insured bank deposits) can provide direct utility by diminishing the fear or worry we experience from the uncertainty that we won’t be able to meet our liabilities or obligations in the future.

One can compare the direct utility that money delivers with the utility provided by a fire alarm: imagine you buy a fire alarm and install it correctly in your home. You may never have a fire in your whole life, and the fire alarm may never in fact go off. Yet the ownership of a working fire alarm nevertheless provides direct utility: it gives you satisfaction or pleasure in diminishing your fears that a fire may occur at night in your house without your knowing or waking up in time. Money’s direct utility works in the same way: it diminishes our fear about the future. Those fears could be manifold: unexpected income loss or money problems from losing our job, accident, sickness, or unexpected future liabilities or obligations.

One of the consequences of recognising that money has direct utility is that the regression theorem – touted by Austrians as the great achievement of their hero Mises (2009 [1953]: 108–111, 121) – becomes entirely otiose.

The whole assumption underlying the regression theorem – that money only has indirect utility – is flawed.

According to Mises, the objective exchange-value of money is “popularly called its purchasing power” (Mises 2009 [1953]: 97). Money is held by people in cash balances, but not to be consumed: money (supposedly) has no use in itself, but is held because of its past exchange value, so that it may be exchanged for goods (Rothbard 2011: 692). What is the cause of the immediate future purchasing power of money? Mises held that the solution is to look at the purchasing power of money in the immediate past (Mises 1998 [1949]: 405). However, there is a problem with this: it appears that the indirect utility of money (by means of its purchasing power) simply depends on its utility (see Graziani 2003: 7–9). The regression theorem was intended to break this circularity.

But the problem that the regression theorem tries to solve – the alleged circularity involved – is nothing but a pseudo-problem. It follows that the regression theorem is essentially pointless and worthless.

What is strange here is that one can find some Austrians who admit that money can yield direct utility (cf. Hutt 1956, although still not exactly an example of this). In particular, one can refer to this talk by Hans-Hermann Hoppe:
Hans-Hermann Hoppe, “‘The Yield from Money Held’ Reconsidered,” Mises Daily, 14 May, 2009,
At one point, Hoppe argues as follows:
“Because money can be employed for the instant satisfaction of the widest range of possible needs, it provides its owner with the best humanly possible protection against uncertainty. In holding money, its owner gains in the satisfaction of being able to meet instantly, as they unpredictably arise, the widest range of future contingencies. The investment in cash balances is an investment contra the (subjectively felt) aversion to uncertainty. A larger cash balance brings more relief from uncertainty aversion. .... The marginal utility of the added cash is higher than (ranks above) the marginal utility of the nonmoney goods sold or unbought.”
Hans-Hermann Hoppe, “‘The Yield from Money Held’ Reconsidered,” Mises Daily, 14 May, 2009,
This can only mean money can provide direct utility as a protection/security against uncertainty.

But Hoppe never thinks about where that leaves the regression theorem.

If it is admitted that money can yield direct utility, then Mises’s whole purpose in thinking up the regression theorem was a waste of time: the imagined circularity he was attempting to solve was an illusion.

I will end by posting the video of Hans-Hermann Hoppe’s talk “‘The Yield from Money Held’ Reconsidered” (delivered at the Prague Conference on Political Economy, 24 April, 2009), with some points following.

First, Hoppe does not refute the quotation of Keynes he cites early in his lecture. For the assertion that the failure to spend money today is likely to diminish consumption and capital goods investments (that is, the increased holding of money is “unproductive” in this sense) is not refuted by invoking the direct utility of money. For the individual, there is no doubt direct utility to be had by holding extra money from the fear of an uncertain future: as Hoppe argues, it can be “productive of human welfare” in this individual or microeconomic sense. But at the aggregate level the effects on production and employment are deleterious. It is the failure to separate micro from macroeconomic effects that destroys Hoppe’s arguments.

And, by the time we get to the end of the talk, Hoppe has long since moved into a la-la land of pure fantasy:
“The situation does not change if there is a general increase in the demand for money, i.e., if all or most people try to increase their cash holdings, in response to heightened uncertainty. With the total quantity of money given, the average size of cash holdings cannot increase, of course. Nor is the total quantity of producer and consumer goods that make up the physical production structure affected by a general increase in the demand for money. It remains unchanged.

In generally striving to increase the size of their cash holdings, however, the money prices of nonmoney goods will be bid down, and the purchasing power per unit money will correspondingly rise. Thus, the (increased) demand for and the (given) supply of money are equilibrated again, but at a higher purchasing power per unit money and lower prices of nonmoney goods.” Hans-Hermann Hoppe, “‘The Yield from Money Held’ Reconsidered,” Mises Daily, 14 May, 2009,
In the real world, of course, prices and wages are generally inflexible, or not flexible to a high degree. Exceptions occur in times of extreme economic crisis when deflation and wage deflation can happen. But, in this case, debt deflation will cripple an economy and equilibrating forces will not work.

However, in pointing to the direct utility of money, Hoppe inadvertently and in a paradoxical way actually confirms Keynesian theory in this lecture.


Graziani, A. 2003. The Monetary Theory of Production. Cambridge University Press, Cambridge.

Hutt, William H. 1956. “The Yield from Money Held,” in M. Sennholz (ed.), Freedom and Free Enterprise: Essays in Honor of Ludwig von Mises. Van Nostrand, Chicago. 196–216.

Mises, L. 1998 [1949]. Human Action: A Treatise on Economics. Ludwig von Mises Institute, Auburn, Ala.

Mises, L. von, 2009 [1953]. The Theory of Money and Credit (trans. J. E. Batson). Mises Institute, Auburn, Ala.

Rothbard, M. N. 2011. Economic Controversies. Ludwig von Mises Institute, Auburn, Ala.

Friday, January 13, 2012

Mises’s Regression Theorem: A Critique

Ludwig von Mises held that money can only have an indirect utility: money has indirect utility by its ability to purchase goods and services. That is to say, the indirect utility of money is derived from the utility of any commodity or set of possible commodities you can purchase with it. Mises stated this in the following way:
“In the case of money, subjective use-value and subjective exchange value coincide. Both are derived from objective exchange-value, for money has no utility other than that arising from the possibility of obtaining other economic goods in exchange for it. It is impossible to conceive of any function of money, qua money, that can be separated from the fact of its objective exchange-value. As far as the use value of a commodity is concerned, it is immaterial whether the commodity also has exchange-value or not; but for money to have use-value, the existence of exchange-value is essential. This peculiarity of the value of money can also be expressed by saying that, as far as the individual is concerned, money has no use-value at all, but only subjective exchange-value.” (Mises 2009 [1953]: 97–98).

“The price of money, like other prices, is determined in the last resort by the subjective valuations of buyers and sellers. But, as has been said already, the subjective use-value of money, which coincides with its subjective exchange-value, is nothing but the anticipated use-value of the things that are to be bought with it. The subjective value of money must be measured by the marginal utility of the goods for which the money can be exchanged. It follows that a valuation of money is possible only on the assumption that the money has a certain objective exchange-value.” (Mises 2009 [1953]: 108–109).
To clarify, according to Mises, the objective exchange-value of money is “popularly called its purchasing power” (Mises 2009 [1953]: 97). The problem, then, for Mises was to account for the “origin of the objective exchange-value of money” (Mises 2009 [1953]: 123). Money is held by people in cash balances, but not to be consumed: money (supposedly) has no use in itself, but is held because of its past exchange value, so that it may be exchanged for goods (Rothbard 2011: 692). What is the cause of the immediate future purchasing power of money? Mises held that the solution is to look at the purchasing power of money in the immediate past (Mises 1998 [1949]: 405).

However, there is a problem with this: it appears that the indirect utility of money (by means of its purchasing power) depends on its utility (see Graziani 2003: 7–9; Schumpeter 1954: 1086–1091).

The Regression Theorem was intended to break this circularity (Mises 2009 [1953]: 97–123; Mises 1998 [1949]: 405–413; Rothbard 2009: 60–61; Rothbard 2011: 692–695). Mises states:
“The objective exchange-value of money which rules in the market to-day is derived from yesterday’s under the influence of the subjective valuations of the individuals frequenting the market, just as yesterday’s in its turn was derived under the influence of subjective valuations from the objective exchange-value possessed by the money the day before yesterday. If in this way we continually go farther and farther back we must eventually arrive at a point where we no longer find any component in the objective exchange-value of money that arises from valuations based on the function of money as a common medium of exchange; where the value of money is nothing other than the value of an object that is useful in some other way than as money. But this point is not merely an instrumental concept of theory; it is an actual phenomenon of economic history, making its appearance at the moment when indirect exchange begins.

Before it was usual to acquire goods in the market, not for personal consumption, but simply in order to exchange them again for the goods that were really wanted, each individual commodity was only accredited with that value given by the subjective valuations based on its direct utility. It was not until it became customary to acquire certain goods merely in order to use them as media of exchange that people began to esteem them more highly than before, on account of this possibility of using them in indirect exchange. The individual valued them in the first place because they were useful in the ordinary sense, and then additionally because they could be used as media of exchange.” (Mises 2009 [1953]: 121).
Money’s purchasing power today is derived from its purchasing power yesterday and so on back to a time when money originally had direct utility as a commodity. The Regression Theorem was also used to explain why fiduciary media or fiat money have and maintain value:
“If the objective exchange-value of money must always be linked with a pre-existing market exchange-ratio between money and other economic goods (since otherwise individuals would not be in a position to estimate the value of the money), it follows that an object cannot be used as money unless, at the moment when its use as money begins, it already possesses an objective exchange-value based on some other use. This provides both a refutation of those theories which derive the origin of money from a general agreement to impute fictitious value to things intrinsically valueless and a confirmation of Menger’s hypothesis concerning the origin of the use of money.

This link with a pre-existing exchange-value is necessary not only for commodity money, but equally for credit money and fiat money. No fiat money could ever come into existence if it did not satisfy this condition. Let us suppose that, among those ancient and modern kinds of money about which it may be doubtful whether they should be reckoned as credit money or fiat money, there have actually been representatives of pure fiat money. Such money must have come into existence in one of two ways. It may have come into existence because money-substitutes already in circulation, i.e., claims payable in money on demand, were deprived of their character as claims, and yet still used in commerce as media of exchange. In this case, the starting-point for their valuation lay in the objective exchange-value that they had at the moment when they were deprived of their character as claims. The other possible case is that in which coins that once circulated as commodity-money are transformed into fiat money by cessation of free coinage (either because there was no further minting at all, or because minting was continued only on behalf of the Treasury), no obligation of conversion being de jure or de facto assumed by anybody, and nobody having any grounds for hoping that such an obligation ever would be assumed by anybody. Here the starting-point for the valuation lies in the objective exchange-value of the coins at the time of the cessation of free coinage. Before an economic good begins to function as money it must already possess exchange-value based on some other cause than its monetary function. But money that already functions as such may remain valuable even when the original source of its exchange-value has ceased to exist. Its value then is based entirely on its function as common medium of exchange.” (Mises 2009 [1953]: 110–111).
Thus a “money thing” must emerge from its original exchange value in direct exchanges, but then it comes to be used as a medium of exchange in indirect exchanges. The process Mises imagines is as follows:
commodity with objective exchange-value traded in direct barter exchange > commodity used as a medium of exchange for indirect exchanges > emergence of money.
The essence of Mises’s theory was summed up by Sir John Hicks as the idea that “money is a ghost of gold – because, so it appeared, money as such has no marginal utility” (Hicks 1935: 2).

But there is a severe flaw underlying Mises’s whole intellectual program in producing his Regression Theorem: the truth of the assumption that money only has indirect utility.

The view that money only has utility through its exchange value is also held by neoclassicals. As the American neoclassical F. W. Taussig argued,
“[t]he phrase “marginal utility of money” must … be used with caution. Money has utility in a different way from other things. It is valued not because it serves in itself to satisfy wants, but as a medium of exchange, having purchasing power over other things. Gold jewelry is subject to the law of diminishing utility precisely as other things are. But gold coin—money—is subject to it only in the sense that an individual buys first the things he prizes most, and then other things in the order of their less utility” (Taussig 1911: 124).
This idea held by Austrians and neoclassicals should be rejected. To begin with, the neoclassical idea that money is neutral is false: the economy cannot be modelled as a mere barter system, as if money is only a numeraire.

Subjective value (or “economic value”) consists in the utility (satisfaction or pleasure) derived by an economic agent from a good. Fiat money may not be a good in the sense of a producible commodity, but money and even financial assets do have utility in themselves, because they have a store of value function in an uncertain world: holding or possessing them allows security against uncertainty, which gives one the ability to meet future uncertain obligations and needs.

Money can even be hoarded directly by some people for the pleasure they might derive from holding large amounts, such social prestige associated with the status of being rich. The existence of the practice of hoarding of money, particularly in times before the widespread use of fractional reserve banking, cannot be doubted either. In a world where we face pure uncertainty in the Knightian sense, cash balances have direct utility (for the history of the idea of preference for liquidity in an uncertain world, see Ritzmann 1999).

But money as the most liquid asset has greater utility than most financial assets, except perhaps the debt instruments we call fractional reserve demand deposits or transactions accounts. Possession of money gives direct satisfaction in providing protection against uncertainty, and allows one to discharge expected and unexpected future liabilities or obligations. Hence money provides a liquid balance that provides the satisfaction of protection against uncertainty. The economist Karl Schlesinger (1914) was an early proponent of this view, and he argued that money held as cash balances can have direct utility, in addition to the indirect utility money can have from its purchasing power (Graziani 2003: 10). John Hicks (1980 [1933]: 528) also appears to have come close to the same view in a 1933 paper.(1) (see Leijonhufvud 2000: 97–98 for an analysis of the wider issues Hicks raised relating to general equilibrium theory).

Money, then, can yield direct utility (Graziani 2003: 11). Post Keynesian economics also takes the view that money (even fiat money) has utility:
“In an uncertain world, the possession of money and other nonproducible liquid assets provides utility by protecting the holder from fear of being unable to meet future liabilities” (Davidson 2003: 236).
Where does this leave Mises’s Regression theorem? The conclusions to be drawn are as follows:
(1) The search for a solution to the alleged circularity of the indirect utility of money depends on rejecting the view that money has direct utility. But money has direct utility, and the whole problem of the alleged circularity that Mises tried to solve is an illusion. It follows that the Regression Theorem is essentially pointless and worthless.

The whole assumption underlying the Regression Theorem is flawed in that the expected future value of money or liquid financial assets can provide direct utility to the holder of money or that asset (for other critiques of Mises’s Regression Theorem, see Patinkin 1965, whose critique already raised the possibility that Mises’s assumption of only an indirect utility for money was wrong; Zazzaro 2003: 237, n. 18; Timberlake 1987).

(2) The regression theorem also commits a non sequitur in ignoring the role that money can have in discharging future obligations such as taxes, and the theoretical possibility and empirical reality that money could arise by governments imposing tax obligations on a community, by taking goods in kind, paying wages for labour, or issuing debt, and giving in return a token or money thing, and demanding that taxes be discharged in that particular money thing (e.g., a hazelwood tallystick, paper money, etc.). Money can then acquire purchasing power through people’s need to acquire it to pay taxes: the purchasing power of money would then develop as related to the value of goods that can also be used to extinguish taxes. The direct utility that money provides would be the ability to meet future tax obligations.

I suspect that real world examples of the introduction of paper money or fiat money provide empirical evidence that refutes the Regression Theorem, including the following:
(1) The introduction of paper money under some dynasties in medieval China (although it is possible that these paper monies were originally convertible into specie; see Glahn 1996 for a starting point in the research literature).

(2) Some of the fiat monies of the British North American colonies in the 18th century.

(3) The issue and use of hazelwood tally sticks by European medieval states as debt money issued to state creditors, which could be used as a medium of exchange and to discharge tax obligations.
A further investigation of these historical examples is clearly in order, though I will not do it here.

In ancient Mesopotamia, money appears to arise first as a unit of account developed from weight units of silver and barley by temple and palace institutions, which was used to calculate the value of taxes, rents and debt, in real goods. Payment could be made not just in silver or barley, but in goods through administered prices denominated in the temple/state unit of account. The use of the unit of account and the medium of exchange was then imposed on the community.

(1) Strictly speaking, Hicks did not appear to regard his demand for cash balances owing to future uncertainty as providing “direct” utility: “ … people always demand money as money – not because it has direct utility to them, but because it is to be used in the making of future payments. In addition to the demand for money as a commodity, there is a demand that arises directly out of ignorance of the future” (Hicks 1980 [1933]: 528).


Anderson, B. M. 1917. The Value of Money, Macmillan, New York.

Davidson, P. 2003. “Keynes’ General Theory,” in J. E. King (ed.), Elgar Companion to Post Keynesian Economics, Edward Elgar Publishing, Cheltenham, UK and Northampton, MA. 229–237.

Ellis, H. S. 1934. German Monetary Theory, 1905–1933, Harvard University Press, Cambridge, Mass.

Glahn, R. von. 1996. Fountain of Fortune: Money and Monetary Policy in China, 1000–1700, University of California Press, Berkeley, Calif. and London.

Graziani, A. 2003. The Monetary Theory of Production, Cambridge University Press, Cambridge.

Hicks, J. R. 1933. “Gleichgewicht und Konjunktur,” Zeitschrift fur Nationalokonomie 4: 441–45.

Hicks, J. R. 1935. “A Suggestion for Simplifying the Theory of Money,” Economica n.s. 2.5: 1–19.

Hicks, J. R. 1980 [1933]. “Equilibrium and the Cycle,” Economic Inquiry 18.4: 523–534.

Hicks, J. R. 1982. Money, Interest and Wages (vol. 2), Blackwell, Oxford.

Huerta de Soto, J. 2006. Money, Bank Credit and Economic Cycles (trans. M. A. Stroup), Ludwig von Mises Institute, Auburn, Ala.

Innes, A. M. 1913. “What is Money?,” Banking Law Journal May: 377–408.

Leijonhufvud, Axel. 2000. Macroeconomic Instability and Coordination: Selected Essays of Axel Leijonhufvud, Edward Elgar, Cheltenham.

Mises, L. 1998 [1949]. Human Action: A Treatise on Economics, Ludwig von Mises Institute, Auburn, Ala.

Mises, L. von, 2009 [1953]. The Theory of Money and Credit (trans. J. E. Batson), Mises Institute, Auburn, Ala.

Murphy, R. P. and A. Gabriel, 2008. Study Guide to Human Action: A Guide Tutorial of Ludwig von Mises’s Classic Work, Ludwig von Mises Institute, Auburn, Ala.

Patinkin, Don. 1956. Money, Interest, and Prices: An Integration of Monetary and Value Theory, Row, Peterson, Evanston, Ill.

Patinkin, Don. 1965. Money, Interest, and Prices: An Integration of Monetary and Value Theory (2nd edn.), Harper & Row, New York.

Ritzmann, F. 1999. “Money, A Substitute for Confidence? Vaughan to Keynes and Beyond,” The American Journal of Economics and Sociology 58. 2: 167–192.

Rothbard, M. N. 1988. “Timberlake on the Austrian Theory of Money: A Comment,” Review of Austrian Economics 2: 179–187.

Rothbard, M. N. 2009. Man, Economy, and State: A Treatise on Economic Principles (2nd edn.), Ludwig von Mises Institute, Auburn, Ala.

Rothbard, M. N. 2009. The Essential von Mises, Ludwig von Mises Institute, Auburn, Alabama.

Rothbard, M. N. 2011. Economic Controversies, Ludwig von Mises Institute, Auburn, Ala.

Schlesinger, Karl. 1914. Theorie der Geld- und Kreditwirtschaft, Duncker & Humblot, Munich.

Schumpeter, J. A. 1954. History of Economic Analysis, Allen & Unwin, London and New York.

Taussig, F. W. 1911. Principles of Economics, Volume 1. Macmillan Company, New York.

Timberlake, R. H. 1987. “A Critique of Monetarist and Austrian Doctrines on the Utility and Value of Money,” Review of Austrian Economics 1: 81–96.

Zazzaro, A. 2003. “How Heterodox is the Heterodoxy of Monetary Circuit Theory? The Nature of Money and the Microeconomics of the Circuit,” in L.-P. Rochon and S. Rossi (eds), Modern Theories of Money: The Nature and Role of Money in Capitalist Economies,Edward Elgar Pub, Cheltenham. 219–245.