[This is a subpost of Bitcoin, Cryptocurrencies and Blockchains.]
[Much of this post has since been published as part of ‘Cryptocurrencies and Digital Fiat Currencies‘]

I begin by reinterpreting some concepts from cryptocurrencies in terms of money in general. I then discuss how widespread their current use is. This includes a broader discussion of fiat-digital-money, as well as the potential implications for monetary policy. I end with a short description of the views of Economists on their future.

Many technical aspects of cryptocurrencies can be reinterpreted in terms of money. Forking the blockchain acts similarly to debasing of money. It trims a fraction of the value off from the forked currency. A historical example of devaluing a currency by trimming off a fraction of the value is Kipper und Wipper in 17th century Germany, when silver coins were debased by mixing them with other metals such as lead.

Double-spending, is something like the digital alternative to counterfeiting. But also very different. Counterfeiting involves producing fake copies of real money, and the difficulty is to make the fake notes look real. For modern currencies this takes a professional group of forgers, whether prisoners-of-war in Germany, or a team in North Korea. For a digital-currency double-spending is subtly different. Creating fake notes is nigh impossible thanks to cryptography, however the same real coin can simply be duplicated and sent to two people, and only one of these two coins will end up existing on the recorded transactions record, the other simply disappears.

Present Importance and Use of Cryptocurrencies

There are two main ways measure the importance of cryptocurrencies. The first is as the total value of all coins. In April 2018 the total market value of all bitcoin was slightly over US$100 billion. By contrast the total value of all US dollar currency is around US$160,000 billion, or US$1.6 trillion. Broader measures of money such as M1 which includes bank deposits, or M2 which also includes savings accounts are US$3.7 trillion and US$14 trillion, respectively. Adding just Euro currency of which there is some 1.2 trillion it is soon clear that Bitcoin, or even all cryptocurrencies combined, are tiny by comparison.

The present importance of cryptocurrencies could also be measured by the volume of transactions, relative to other means of payment. In April 2018 the average number of transactions per second being processed by the three largest cryptocurrencies are 3-4 by Bitcoin, 15 by Ethereum, and 13 by Ripple. By contrast Paypal averages 200 and Visa averages nearer 1700 transactions per second. So this measure also shows role of cryptocurrencies in transactions to be tiny in comparison to other methods.

In fact using cryptocurrencies as means of payment to conduct transactions, at least for Bitcoin, does not even appear to be the main interest. From 2011 to 2013 new Bitcoin users appear to correlate with increased activity on Bitcoin exchanges, which makes sense if they wish to use Bitcoin as a speculative asset, but does not correlate with increased transactions on the Bitcoin blockchain, which is what we would expect to see if they were used to transact (Glaser et al., 2014).

For the present cryptocurrencies remain a minor niche, albeit one displaying a sharply upward trend.

Will Cryptocurrencies take over?

Monetary theory suggests that a large number of coexisting cryptocurrencies together with a stable value of money is a possible outcome. But that it is just one of many such possible outcomes in a world of competing cryptocurrencies. This modern monetary theory (a.k.a. search-money theory) also suggests that a government issued fiat-currency would be the dominant currency, provided the government controls inflation. Whether or not the government issued fiat-currency allowed for the coexistence of cryptocurrencies would depend on government policy. This theory leads to the conclusion that cryptocurrencies are unable to compete with a ‘good’ fiat-currency, but that they would displace a ‘bad’ fiat-currency—bad meaning one associated with high inflation (Fernandez-Villaverde and Sanches, 2018). This suggests that cryptocurrencies are likely to remain a niche in countries with low and stable inflation, such as the US and the Eurozone in recent decades. But countries such as Venezuela, where inflation in the Venezuelan bolivar is forecast to exceed 10,000% in 2018, could see their currency pushed out and replaced by cryptocurrencies.

Notice that this theory formalizes some ideas of Hayek (1976), except that Hayek argued in favour of currency issued by private Banks precisely because he didn’t believe that the relatively stable value of fiat-currencies achieved by most developed countries over the past thirty years were possible. Unlike Hayek though who glossed over the question of the incentives of private Banks to issue such currencies¹, cryptocurrencies have tackled the incentives of the currency ‘issuers’ head on.

Fiat-currencies also have two major home-ground advantages. They are legal tender: they must be accepted as payment for the settlement of contracts. And the government both requires them to be used to pay taxes and uses them to make payments itself. Governments could of course change the laws to remove these advantages by allowing cryptocurrencies to be used alongside fiat-currency as legal tender and for paying taxes—in fact current US law allows contracts to written so that the settlement has to be in Gold, rather than US dollars— but for the present that seems unlikely. Other than the European Union countries typically tax cryptocurrencies as assets, not currency (Kumar & Smith, 2018).

That cryptocurrencies seem unlikely to take over is far from saying they won’t play a role alongside more traditional forms of money. Among their strengths they appear a good way to transfer money internationally, since you can simply use say New Zealand dollars to buy cryptocurrency and then sell these for US dollars. This is useful both in keeping fees low for converting currency, and in some countries for avoiding limits on foreign exchange.

Fiat-Digital-Money (or Fiat-Crypto)

So decentralized cryptocurrencies seem unlikely to replace existing money. But why not have the fiat-money itself be a cryptocurrency?, an idea variously called fiat-crypto, fiat-digital-money, or ‘digital Central Bank money’. Two cases are made: the first envisions it replacing cash alone and I will call this fiat-digital-currency, the second as becoming used for all payments meaning it would also replace bank transfers, deposit accounts, and debit cards and I will call this fiat-digital-money. The case for replacing cash with fiat-digital-currency is made on two different grounds: replacing cash with fiat-digital-currency is argued for either to reduce crime, bribery and fraud (which often use cash), or because it is cheaper to operate. The wider use of a fiat-digital-money that could be used for all payments tends to be more about reducing the role of private banks. A fiat-digital-currency intended to replace just cash could pay no interest, but to be able to replace other forms of payment fiat-digital-money would have to pay interest to able make the alternatives unattractive.

One rationale for replacing cash with fiat-digital-currency is to reduce crime, especially money laundering and fraud, and to reduce the size of the informal sector (cash-in-hand work that avoids taxes and regulation). For this to work the fiat-digital-currency accounts could not be anonymous, which clearly opens up large concerns about privacy. Even if done, unless the use of alternatives such as cryptocurrencies are all banned it seems unlikely this would achieve much in terms of reducing crime. Another rationale is that a fiat-digital-currency might be cheaper to operate than existing fiat-currency: partly because the cost of moving cash around is higher than digital payments, partly because some small denomination coins actually cost more to mint than they are worth. The potential cost savings of fiat-digital-currency appears to be a substantial part of the interest in Sweden where declining use of cash has made handling it more expensive relative to transactions volume.² A final rationale for replacing cash with fiat-digital-currency is to make it easier to set negative interest rates; although this could likely be achieved simply using fiat-digital-currency alongside cash (Agarwal & Kimball, 2015).

Using narrow fiat-digital-currency to eliminate cash entirely would require some major changes in the ‘infrastructure’ of money. Currently only private Banks have access to accounts at the Central Bank. The use of cash, while involving some Government role in its creation and distribution, largely involves the private sector processing cash transactions. The Central Bank typically plays little role in digital transactions (including credit and debit cards) between Merchants and Banks, but does play a large role in transactions between Banks in many countries. A fiat-digital-currency would (presumably) involve opening up to allow anyone to hold an account at the Central Bank, and having the Central Bank directly involved in all transactions using the fiat-digital-money.³ For private money not to be displaced requires the fiat-digital-currency to be inferior to other means of payment, and paying zero-interest would (be one way to) ensure this.⁴

The approach of introducing a fiat-digital-currency to replace only cash would leave the existing bank system in place. Under the current system Central Banks issue a certain amount of cash, a public money the supply of which they can precisely control. Private banks then issue private money. They could issue this private money with no backing whatsoever, that is with no guarantee you could exchange ‘mybank dollars’ for dollars, this would look a lot like cryptocurrencies in the sense that the price (or exchange rate) between ‘mybank dollars’ and dollars would fluctuate as do the prices of cryptocurrencies. Instead the private money issued by private banks is pegged to the dollar: you can exchange your bank deposit for cash at any time at a fixed rate of one-for-one. To make this exchange credible private banks need to hold some cash, and this system is called fractional reserve; private banks hold cash reserves equal to a fraction of the private money they issue as bank deposits. This system of fractional-reserves allows private banks to alter the amount of money in response to changes in demand for money. However it also means that a given bank’s private-money is only of value if people believe that it has enough cash reserves to satisfy the demand for them. If this ceases to be the case then people will no longer value the private money and there is a run on the bank as people attempt to convert them into cash reserves before those run out.⁵

With the Great Financial Crisis of 2007 many people lost faith in the ability of private banks to maintain sufficient reserves. One bank, Northern Rock, saw a classic bank run. Many other banks saw the modern equivalent, a run in the overnight repo market, the equivalent of bank deposit accounts for large corporations. Fractional-reserves is what makes banks vulnerable to runs. So a move to requiring banks to hold full-reserves would lead to a system in which bank runs were impossible.⁶ But under a full-reserve system private banks play no role in determining the amount of money, the difference between the bank deposits and cash would just be that one is digital and the other physical cash. In that case why not have the Central Bank issue digital money directly? This is precisely the idea of a fiat-digital-money that replaces all payments.

Both the advantages and disadvantages of a fiat-digital-money that replaces all payments are precisely that it would eliminate the Banking system. At least in its current form the Banking system is almost entirely funded by bank deposits (and repos, essentially large bank deposits for big corporations) while with a fiat-digital-money people would instead hold these at the Central Bank. The advantages and disadvantages of this fall in three main categories, risk, expense, and monetary policy.

In the risk category, the advantages of fiat-digital-money include eliminating the possibility of Bank runs and some associated financial panics. But the Central Bank would take on these huge risks that are currently borne by the private Banks, making the taxpayer directly liable for all these risks. In 2015 Switzerland abandoned a currency ceiling against the Euro precisely because it was exposing them, via the Central Bank balance sheet, to huge financial risks around the value of assets and the exchange rate.

Providing money is hugely expensive. Presently Central Banks pay for minting currency by seignorage, and leave the private sector to pay most of the expense for transporting and securing cash, while private Banks pay for the expense of bank-deposits by lending in the form of credit-cards, mortgages and commercial loans. A fiat-digital-money would leave the Central Bank paying for all of this, and because crowding-out other forms of payment would require the Central Bank to pay interest on deposits the expenses would be huge.⁷ While one advantage of fiat-digital-money is precisely that it would be cheaper than printing, transporting, and securing cash all of the advantages of this were already present in the narrower vision of fiat-digital-currency. Central Banks are lightly profitable for Governments, the US Federal Reserve makes an average of $90 billion per year from seignorage (inflation is essentially a tax on money holdings), even after covering all its running costs. Taking on all the expense of providing fiat-digital-money that displaced private Bank deposits would instead turn Central Banking into a huge expense for Government. This would lead to the temptation to pay for this by printing money, leading to high inflation; clearly a bad idea. Or to pay for it the same way as private Banks, by lending out the (otherwise idle) funds. This would leave the Government, via the Central Bank, directly exposed to all the kinds of risks currently borne by the private Banks. The experience of the Spanish public banks (las Cajas) in the Great Recession —where they turned out to have taken more risks and ended up losing much more money than the private banks, requiring widespread bailouts— suggests this is unlikely to be a good idea. This returns us to the idea that fiat-digital-money would mean the Central Bank taking on risks that are currently borne by the private Banks, making the taxpayer directly liable for all these risks. This is undesirable.

Another related risk is that with the Central Bank responsible for the entire money supply, rather than just currency, there would be less constraint on the temptation for the Government to fund itself by seignorage, that is, by printing money. This was the eventual outcome in the USSR where Gosbank (the Soviet central bank) issued the pen-and-paper equivalent of fiat-digital-money in the form of both cash and non-cash rubles. While some countries such as the US and Eurozone have capably resisted this temptation in recent times, it is not clear how they would fare with the higher temptations of a fiat-digitial-money. For other countries like Argentina, Venezuela and Zimbabwe, even the present temptations of funding themselves by printing money have proven much to large to resist. Argentina has experienced decades of substantial inflation as a result, and both Venezuela and Zimbabwe have recent experience with hyperinflation.

The other major issue is the value of money and monetary policy. As long as the supply of fiat-digital-money grew at the same rate as the real economy over periods of a decade the quantity theory of money tells us that the value of money would be stable over longer periods of time. In the short-run Central Banks would have a choice. They could allow the supply fiat-digital-money to expand and contract in response to changes in demand in order to maintain price stability (against a target basket of goods and services). They could keep the supply of fiat-digital-money fixed, and allow prices (the inverse of the value of money) to fluctuate in response to changes in demand; this would be a brave new world for money. Or they could allow the supply to partly respond, or even over-respond. This would lead to some fluctuation in prices (the inverse of the value of money) and in effect would be much the same as current active monetary policy does by changing interest rates. What exactly Central Banks would choose is unclear, but it seems likely they would continue to perform active monetary policy. The Central Bank could even set an interest rate policy for the return on fiat-digital-money and let the supply fluctuate to meet demand. Whether in practice this would leave them with more or less control over the economy, and whether the outcomes for economic performance would be better or worse is hard to tell.

There is another possibility, which is that a country introduces a fiat-digital-currency alongside existing cash. This would allow the advantages of lower cost and more convenient transactions from digital-money to coexist with the privacy advantages of cash. Many countries seem to be striking a balance along these lines.

Summing up with my own opinion. Fiat-digital-currency is a good idea, being cheaper and a better method of payment than cash, but should be inferior to private bank deposits by design. The alternative would see it displace bank deposits and become a fiat-digital-money, leading either to huge inefficiencies (if everyone left their money idle in accounts at the Central Bank) or else to the Central Bank taking on large expenses and financial risks currently borne by private Banks and the financial system. This means fiat-digital-currency should either have zero nominal return, or perhaps even slightly negative returns. Given that the real return on cash is roughly minus the inflation rate (so around -2% on average in many OECD countries) slightly negative real returns on fiat-digital-currency are not as odd as they may first sound. A fiat-digital-currency should be seen as a complement to cash, rather than replacing it entirely.

Economists Opine on Cryptocurrencies

I finish with the current views of Economists on what should be done about cryptocurrencies. They range from full fiat-digital-money, through hands-off, to seeing them as pointless or even an outright threat needing heavy regulation. The most common view is probably be that they should be treated with a mix of curiosity and caution, seeing them as presently ineffectual but potentially of great importance in the future. Whether for good or bad remains to be seen.

My own view is that the current generation of cryptocurrencies is more like digital-gold than digital-money. They may come to play an important role in international money transfers as you can change from exchange from one currency, to a cryptocurrency, to another currency, at low cost. The main regulatory concerns relate to money laundering, and their possible use to fund illegal activities. Future generations of cryptocurrencies will solve the difficulties relating to means-of-transaction, but whether they can deliver a stable store-of-value remains to be seen.

 

---

Remark: Fiat-Crypto would be a digital money, but would be a centralized rather than decentralized currency, with the Central Bank in control. Given this it is not clear why the blockchain would be the preferred database technology that the Central Bank would want use to implement such a digital currency. Hence I prefer to use the term Fiat-digital-money.

Remark: Cash in New Zealand.

Remark: The search-money of theory is based on Kiyotaki & Wright (1993) – A Search-Theoretic Approach to Monetary Economics and Lagos & Wright (2005) – A Unified Framework for Monetary Theory and Policy Analysis.

---

Footnotes

1. Hayek’s view was that competition would lead Banks to issue private currencies of a stable value, and modern monetary theory suggests this is one possible outcome (one of many competitive equilibria, and one which exists only under certain assumptions about beliefs and the cost of emitting private currencies; Martin & Schreft, 2006). But an analysis of the effects of competition on the incentives of the Banks suggest this outcome not efficient (Monnet & Sanches, 2015) since people require currency for transactions purposes the efficient outcome requires them to be able to make some return on this currency, and competition means that the Banks are not able to make enough profits elsewhere to pay such a return on the money they issue. The logic is similar to the Friedman rule, common in much monetary theory, that argues that money should pay a real return to compensate the opportunity cost of not holding other assets. For a fiat-currency this would mean steady deflation, and while theory says this is a possible equilibrium, this concept of efficiency is not achieved in real-world practice. Further, while private monies issued by Banks in an environment of perfect competition are consistent with the existence of a stable value of money equilibrium, they are also subject to the same self-fulfilling inflationary episodes that can occour in economies with just a fiat-currency, even when the Banks are profit-maximizing, long-lived entrepreneurs who care about the future value of their monies (Fernandez-Villaverde & Sanches, 2018WP for private Banks; Lagos & Wright, 2003 for fiat-currency).

2. Sweden’s case illustrates the two-way feedback between the ‘expense’ of transacting in cash and its use. The only place in Sweden where banks can get cash from the Central Bank is at a cash window near Arlanda airport near Stockholm. So the private sector is Sweden has to pay the expense of managing and transporting cash. Previously there had been multiple cash windows around the country, and when these were closed it led to an increase in the frequency of electronic transactions. This has meant cash is now little used and thus even more expensive to process. By contrast Germany and Japan have very low interest rates, and so the opportunity cost of using cash is low (the returns to instead holding money in bank accounts are low) and both countries continue to see widespread use of cash.

3. The alternative of having the Central Bank only let Banks access fiat-digital-currency, and allowing all other transactions in the economy to be made by, .e.g., debit cards looks more like simply eliminating cash, than replacing it with a fiat-digital-currency.

4. In recent times when the zero-lower bound on nominal interest rates was reached even this may not be enough to avoid displacing other means of payment and a negative interest rate on the fiat-digital-currency might be required.

5. I am describing a liquidity crisis. In a solvency crisis the problem is that the Bank no longer has enough assets to pay everyone back; e.g., it might have lent mortgages which have been defaulted on. In a liquidity crisis the Bank has enough assets to pay everyone back eventually, however it does not have enough cash on hand to repay everyone bank right now, and so if they all demand to be repaid now then a bank run occours and the Bank collapses. This view of Banks corresponds to their role in maturity transformation, turning short-maturity debt in the form of bank deposits into long-maturity debt in the form of mortgages and commercial loans.

6. Another closely related, but importantly different, idea is that banks should switch from funding their lending with bank deposits, essentially debt which has a fixed exchange rate to the dollar and so are vulnerable to runs, to funding with equity, which has a fluctuating value to the dollar. This argument is explained in detail in The Bankers New Clothes. Note that this would be in conflict with the idea that maturity transformation (turning home mortgage debt into debt in the form of bank deposits) is a role of Banks.

7. The alternative to paying interest on deposits, namely simply making all alternatives illegal would be hugely distortionary to the economy and likely to lead to large amounts of black-market activity.

---

References

Agarwal and Kimball (2015WP) – Breaking Through the Zero Lower Bound

Carstens (2018 BIS Speech) – Money in the Digital Age: What Role for Central Banks?

Fernandez-Villaverde and Sanches (2018WP) – On the Economics of Digital Currencies

Hayek (1976) – The Denationalisation of Money (and 1978, The Argument Refined)

Glaser, Zimmermann, Haferkorn, Weber, and Siering (2014) – Bitcoin – asset or currency? Revealing users’ hidden intentions
Note: They measure new users by number of people checking Wikipedia page for Bitcoin and Bitcoin Exchange activity based solely on the (public after a hack) data for Mt Gox (which was 85% of the market at the time), transactions on the blockchain are publicly observable.

Kumar and Smith (2018) – Crypto-currencies – An introduction to not-so-funny moneys

Lagos & Wright (2003) – Dynamics, cycles, and sunspot equilibria in ‘genuinely dynamic, fundamentally disaggregative’ models of money

Martin & Schreft (2006) – Currency Competition: A Partial Vindication of Hayek

Monnet (2006) – Private Versus Public Money

Monnet and Sanches (2015) – Private Money and Banking Regulation