Can Currency Competition Work?

There is a new model and NBER paper to come from Jesus Fernandez-Villaverde and Daniel Sanches (pdf, ungated), here is the abstract:

Can competition among privately issued fiat currencies such as Bitcoin or Ethereum work? Only sometimes. To show this, we build a model of competition among privately issued at currencies. We modify the current workhorse of monetary economics, the Lagos-Wright environment, by including entrepreneurs who can issue their own fiat currencies in order to maximize their utility. Otherwise, the model is standard. We show that there exists an equilibrium in which price stability is consistent with competing private monies, but also that there exists a continuum of equilibrium trajectories with the property that the value of private currencies monotonically converges to zero. These latter equilibria disappear, however, when we introduce productive capital. We also investigate the properties of hybrid monetary arrangements with private and government monies, of automata issuing money, and the role of network effects.

I would stress a few points.  First, the world is going to have some form of currency competition whether one likes it or not.  That is already the case today, so these are very real questions, not just thought games for libertarians.

Second, the Bitcoin and broader cryptocurrency experience indicate that the marginal cost of issuing a new private currency is well above zero, contra some of the literature from the 1970s, which viewed the enterprise in terms of printing money and paying only for the additional paper.  Bitcoin can be interpreted as a commodity currency of sorts, where the relevant expenditures are on codebreaking and electricity, rather than digging up gold from the ground.  Once it is focal enough, it might be able to provide some version of rough price stability in terms of its unit.

Third, if your government is halfway legitimate and not broke, its currency is likely to be a dominant winner in these forms of currency competition, especially to the extent that currency is supported by the fiscal authority.  In this sense it is almost impossible to get away from a legitimate or even semi-legitimate government-issued currency.


Paging George Selgin...ex alum of GMU...he knows that when you have competing currencies, even in the 19th C, they coalesce to a common par value, say "one $". See if I'm first...

OT - I like this Lagos-Wright framework, it in some forms assumes money neutrality, which is sound, and does not rely on price stickiness as an assumption: (2010 paper on New Monetarism: “A more obvious illustration of New Monetarists worrying relatively more about the soundness and consistency of economic theories may be the reliance of the entire Keynesian edifice on a foundation of sticky prices, which are not what we would call microfounded, even when—especially when—appeal is made to Calvo (1983) pricing or Mankiw (1985) menu costs. we offer as food for thought two New Monetarist models that speak to the issue. …The other model uses search theory to get nominal rigidities to emerge endogenously, as an outcome rather than an assumption. This model is consistent not just with the broad observation that many prices appear to be sticky, but also the detailed micro evidence discussed by Klenow and Malin (forthcoming) and references therein. Yet it delivers policy prescriptions very different from those of New Keynesians: Money is neutral. “)

...or rather money neutrality is a consequence of the model, which has endogenous nominal rigidities. This is akin to what I wrote upstream in the OP: when you have a bunch of competing currencies, they converge to a common par value, say "one dollar", so that private money from Bank A is worth the same, per unit, as private money from Bank B, as money from the US Treasury, namely, "$1". It's a convenience factor, the same as when in raising chickens the common weight is typically 1 kg per dressed bird (it's what the consumers expect), or 1 liter of soft drink regardless of brand. Nominal rigidities. But that's not the same as 'sticky prices', and the policy prescriptions are different, namely, money is neutral so Fed tinkering will have no effect on the real economy, even short-term.

BTW a trivial aspect of the paper--and one I ignored--is that any currency backed by nothing will go to zero. Tell me something I don't know...

Bitcoin: Sound as a Lopez.

"at" ==> "fiat"

If ransomware attacks that demand payment it bitcoins (like this one became much more widespread, what effect would that have on Bitcoin?

Aren't there monetary theories that maintain that a country's fiat currency ultimately derives its value from the fact that the country demands that taxes be paid using the currency? Perhaps ransomware could serve a similar function for Bitcoin.

I'd understood the value of a fiat currency to loosely represent the total value of the production of the country in the international trading system.

Testable implication: when a fiat money becomes worthless, it must be the case that the country in question has ceased to produce valuable (exportable) goods.

Depends if you are talking about the total amount of a particular fiat currency in circulation, or an individual unit of such.

1920s Germany was producing exportable goods, so the total amount of Marks in the world had some value in other currencies, but due to hyperinflation each individual Mark was worthless, and since anyone holding a currency must by definition hold a certain number of individual currency units, by extension all Marks were worthless.

If the US increased the money supply 100-fold, each unit would decrease in value by 99%, but the value of the total stock would remain the same. Well, not in practice due to its role as reserve currency, etc., but loosely speaking, I think it still applies.

Not generally so. I high rate of inflation tends too reduce the real demand for the depreciating currency; hyperinflation can reduce it to zero.

This reminds me of the Ray Dalio profile published in The New Yorker several years ago in which he defended currency trading (from which he made billions) as providing a public service: an efficient allocation of a scarce resource (money). I mention this because it highlights an aspect of capitalism, namely, that morality, or the common conception of morality, has nothing to do with it. And so it is with the scandal du jour: trillions being diverted to secret bank accounts in tax havens. Is Company A morally complicit if it shifts its production to China knowing that billions will be diverted to what is essentially a criminal enterprise? Or is morality irrelevant if shifting production to China is efficient? It's not often we have the occasion to consider morality in economics. I don't expect this occasion, like all occasions of the past, will last very long.

Trillions in secret bank accounts just end up as negative interest Japanese bonds, etc. Social good accomplished.

Jesus sent me his and Daniel's paper yesterday, so I have only had a chance to glance through it quickly. Some features that I noticed are the following: first, unlike some earlier papers from the same research program, this one is unambiguously about competing fiat monies, as envisioned by Hayek and Benjamin Klein, among others. It doesn't conflate irredeemable private currencies with redeemable banknotes, and therefore doesn't pretend to shed light on the workings of "free banking." That's an improvement in the sense that the other papers appeared to show the free banking couldn't work, when in fact the systems they modeled had little in common with historical free banking systems.

The authors' first consider the case of an economy without productive capital. They find that, although an equilibrium exists in which the rival monies all maintain a stable value, as Hayek predicted would happen, they also find other equilibrium in which the monies lose value steadily. They point out, furthermore, that even the Hayek-type equilibrium isn't efficient compared to one in which currencies appreciate at Friedman's optimum quantity of money rate. Hayek, in contrast, simply assumes that consumers prefer money of constant purchasing power. Even if one takes for granted that stable purchasing power of money is macro-economically desirable, Hayek's assumption makes little sense, since those macroeconomic advantages are far less likely than rate of return considerations are to influence peoples' financial asset preferences. (FWIW, so long as the rate of deflation doesn't exceed the economies rate of productivity growth, it's not bad from a macro point of view, either). Interestingly, when sufficiently productive capital is allowed for, the same model implies a unique equilibrium in which monies issued by rival currency suppliers appreciate at the optimal rate.

Tyler alludes to one troublesome aspect of the analysis, which is that it assumes that private currencies can be produced at zero marginal real cost. That's certainly not the case for Bitcoin and other cryptocurrencies, which (as Tyler also observes) are in this respect more like commodity monies than like textbook "fiat" monies. (I've tried to get people to call them "synthetic" commodity monies rather than "fiat" monies for this reason and also because government compulsion plays no part in their acceptance.) A related and even more troublesome assumption, in my opinion, is that the market for fiat currencies can be regarded as perfectly competitive. That assumption drives many of the papers' results, starting with a zero profit equilibrium condition that in turn implies that all issues maintain a constant money stock, and therefore cannot regulate their currencies' purchasing power by means of quantity adjustments.

The assumption private issuers maintaining a constant nominal quantity of each brand of private currency is certainly not what Hayek had in mind. But then, neither is the assumption of perfect competition. Strictly speaking, such competition requires that suppliers produce identical, that is, indistinguishable, products, which is rather more like legalized counterfeiting, the results of which are easy enough to imagine. Evidently, some kind of product differentiation must be taken for granted if other outcomes are to be at all likely. Hayek and Klein and most others writing on the topic of competing fiat monies took some degree of imperfect or monopolistic competition for granted. An important issue, therefore, is whether the assumption of perfect competition is appropriate and, if not, whether an alternative assumption would lead to substantially different conclusions.

Thanks for that. I got a bit lost in the second paragraph, but soldiered on. That crypto currencies are synthetic commodity money makes a lot of sense. Choices then become the choice of synthesis (hashes etc) and commodity (directly CPU cycles, indirectly electricity).

I think the fact that a backyard-solar Bitcoin mine can't pay says that they've stretched that one too far. Better return farming escargot.

I always thought bitcoin could only really take off if a government began to use it.

Who would do such a thing?

Maybe a place like Ecuador which had to dollarize in the past.

So, if you're forced to use a hard currency, but maybe you don't like the US, you move to bitcoin.

As I understand it the only people who can mine at scale are Chinese who do not pay market (or perhaps any) cost for coal powered electricity. Ecuador would be better off using potatoes or alpaca blanket money. Transactions that do not require a corrupt Chinese partner.

The distributed aspect of bitcoin is the expensive one, and you only need it in low-trust environments.

A semi-trustable government could do an electronic currency that had Bitcoin's transparency features without its distributed features, at significantly lower information processing overhead.

"when we introduce productive capital."

Damn, I want to create wealth without working and that is possible only by extracting rents on movement of money which requires an ever increasing flow of money which creating new monies let's me do!

Creating/maintaining/monitoring movement of money is not "without working." May not be respectable/honorable but it's definitely work ....

Comments for this post are closed