Sunday, April 15, 2018

Critiquing the Carney critique of central bank digital currency

Over on the message board we've been discussing the implications of central bank-issued digital currency, otherwise known as CBDC. One view is that a central bank digital currency would lead to increased financial instability, Bank of England governor Mark Carney being a vocal proponent of this idea. There are a lot of criticisms that can be leveled against central bank digital currency, but the Carney critique is the one that worries me the least. Let's see why. 

First off, let's establish what we mean by digital currency. Imagine that a central bank has discovered a technology that allows it to create an exact digital replica of the banknote. Like banknotes, these digital tokens are anonymous and untraceable. To make use of them, people don't have to register for an account. Rather, the tokens are held independently on one's device, sort of like how paper money is held in one's wallet without requiring any sort of registration with the issuing central bank. This combination of features makes it impossible for the central bank to censor or prevent people from using digital currency, in the same way that the central bank can't stop people from trading paper money among themselves.

Unlike banknotes, which can only be passed face-to-face, digital currency can be transferred instantaneously over the internet. There are no storage and handling costs. $10 million dollars worth of $20 bills takes up a lot of space and is awkward to carry around, but in the digital world that same nominal amount has neither volume nor weight. Lastly, digital currency is cheap to create, requiring only a few keyboard strokes. Cash requires large printing machines, ink, and paper.

Having established what a digital currency is, let's introduce it into the economy. The central bank announces a demonetization of all banknotes and coins, offering $1 of digital currency for each $1 worth of cash. Anyone who want to withdraw money from their bank account will now get digital currency, not banknotes. No one visits ATMs or the bank teller anymore to make a deposit or withdrawal: with an internet-connected device, deposits and withdrawals can be made from bed, the toilet, or while commuting on the bus.

Carney's contention is that the introduction of a digital currency could hurt the banking system:
"...a general purpose CBDC could mean a much greater role for central banks in the financial system. Central banks may find themselves disintermediating commercial banks in normal times and running the risk of destabilising flights to quality in times of stress."
First, let's deal with Carney's normal times critique. The idea here is that by introducing a digital version of the banknote, a significant proportion of existing depositorsthose with chequing and savings accountswill desert their bank because they want to hold sleek and shiny central bank digital currency instead. (Presumably they didn't desert their banks when banknotes were around because cash was bulky and couldn't be transferred instantaneously over a communications network.) By causing a mass draining of depositsi.e. disintermediating commercial banksa new digital currency would impair the ability of banks to make loans, and this would affect the economy in a negative way. 

To show why I don't think the Carney critique holds, we need to investigate one of the important differences between cash/digital currency and bank deposits. When people open bank accounts, what interests them is not just the idea of making payments with those accounts but also maintaining a relationship with the bank in order to benefit from a smorgasbord of other financial services. People with bank accounts are like subscribers to a magazine, they want an ongoing connection.

Those who use cash, on the other hand, would rather just buy the magazine once rather than subscribe to it, orfor another analogyprefer using disposable plastic plates to maintaining a set of their own plates. Cash is a one-time use commodity; once you spend it, any relationship to its issuer is severed. This lack of an ongoing connection provides value to some people. Consider the process of budgeting. By sticking some cash in an envelope dedicated to groceries, another for rent, presents, entertainment, clothing, you can closely monitor your spending over the course of a month. Once the cash is used up, spending stops. With a bank, however, a connection remains even after someone's balance has fallen to zero, spending potentially continuing via overdrafts and credit cards. People who may not trust themselves to stay within their means may therefore prefer the one-time use nature of cash.

So when digital currency replaces cash, I don't anticipate a mass migration from bank accounts to digital currency. Depositors who have already chosen a subscription-based banking solution over a one-time payments solution won't change their minds when the next generation one-time use product is introduced. Which isn't to say that there won't be some sort of migration out of bank deposits and into a new digital currency. Consider upstanding members of society who have always wanted to make anonymous digital payments but haven't had the chance to do so because the only anonymous option theretofore available to themcashwas a physical medium, and so instead they have opted for the inferior option of non-anonymous digital payments services of a bank. This group of anonymity seekers will make the switch. 

But the migration of legitimate anonymity seekers out of bank deposits into digital currency will be counterbalanced by a reverse migration out of cash into bank deposits. Let's think for a moment about who uses cash. Illicit users like criminals and tax evaders are big users, and when cash is demonetized they will all shift into digital currency in order to preserve their anonymity. Likewise, licit users of cash who want to keep using a one-time use payments option will opt for digital currency. The undocumented and those who are too poor to qualify bank accounts will also make the migration into censorship resistant digital cash.

That leaves one major group of cash users unaccounted for: those who use cash not because they like any specific feature that it provides but out of pure force of habit. With cash being cancelled, habitual users will have no choice but to switch into some other payments option. And since deposits are the time-tested option, it is likely that many will move their funds into the banking sector. If this wave of inbound habitual users is greater than the wave of outbound anonymity seekers, then the introduction of a digital currency may actually be lead to an increase in bank intermediation rather than Carney's disintermediation!


So if a digital currency won't affect the banking system during regular times, what about Carney's times of stress criticism? The general criticism here is that during a crisis, households and businesses will desperately shift their deposits into the ultimate risk-free asset: central bank money. Presumably when deposits were only redeemable in banknotes (as is currently the case) and one had to trudge to an ATM to get them, this afforded people time for sober contemplation, thus rendering runs less damaging. But if small depositors can withdraw money from their accounts while in their pajamas, this makes banks more susceptible to sudden shifts in sentiment, goes the Carney critique.     

I don't buy it. Small depositors won't exit banks during a crisis because their money is insured up to $250,000 (in the US). But even in jurisdictions without deposit insurance, I still don't think that shifts into digital currency in times of stress would exceed shifts into banknotes. A bank will quickly run out of banknotes during a panic as it meets client redemption requests, and will have to make arrangements with the central bank to get more cash. Thanks to the logistics of shipping cash, refilling the ATMs and tellers will take time. In the meantime a highly visible lineup will grow in front of the bank, exacerbating the original panic. Now imagine a world with digital currency. In the event of a panic, customer redemption requests will be instantaneously granted by the bank facing the run. But that same speed also works in favor of the bank, since a request to the central bank for a top-up of digital currency could be filled in just a few seconds. Since all depositors gets what they want when they want, no lineups are created. And so the viral nature of the panic is reduced.

But what about large depositors like corporations and the rich who maintain deposits well in excess of deposit insurance ceilings? During a crisis, won't these sophisticated actors be more likely to pull uninsured funds from a bank, which have a small possibility of failure, and put them into risk-free central bank digital currency?

I disagree. In a traditional economy where banknotes circulate, CFOs and the rich don't generally flee into paper money during a crisis, but into short-term t-bills. Paper money and t-bills are government-issued and thus have the same risk profile, t-bills having the advantage of paying positive interest whereas banknotes are barren. The rush out of deposits into t-bills is a digital one, since it only requires a few clicks of the button to effect. Likewise, in an economy where digital currency circulates, CFOs are unlikely to convert deposits into barren digital currency during stress, but will shift into t-bills. The upshot is that banks are not more susceptible to large deposit shifts thanks to the introduction of digital currencythey always were susceptible to digital bank runs thanks to the presence of short-term government debt.

The ability to mitigate shifts out of the banking system during times of stress may be even more potent in a world with digital currency than one without. During a crisis a central bank will generally reduce its main policy interest rate in order to stimulate the economy, short-term market interest rates falling in sympathy. Now, consider an economy with banknotes. Even as short-term rates fall, the interest rate on banknotes stays constant at 0%, the effect being that the relative return on banknotes steadily improves. This only encourages further shifts out of the banking system into cash.

Digital currency updates the cash model by introducing a wonderful new invention: the ability to adjust the interest rate on cash. Now when the central bank reduces its policy rate to offset the weakening economy, it can simultaneously reduce the rate on digital currency. This has the effect of maintaining a constant relative return on currency throughout the crisis. So unlike a banknotes-only world in which the relative return on notes steadily improves as the crisis deepens, thus encouraging disintermediation of the banking sector, a digital currency-only world guards against the sort of return differential that might engender disintermediation.


So the Carney critique, which frets over mass adoption of digital currency, doesn't amount to much, in my view. A better critique of digital currency is the exact opposite: instead of mass adoption, it is very possible that no one (apart from criminals and tax evaders) uses the stuff.

Let's see why digital currency could fail on takeoff. One potential migration pattern I mentioned above involves upstanding members of society who desire anonymous online payments adopting digital currency. But what if there just aren't that many people who care about online privacy? Countries like Sweden, where banknote usage is plummeting, give credence to this concern while surveys of cash users in the eurozone show that anonymity is not terribly important to them:

Another large base of potential digital currency users includes all those who value cash for both its throw-away nature and lack of censorship. But what if these people choose to adopt pre-paid debit or credit cards instead, both of which are open systems that do not obligate users to maintain an ongoing relationship with the issuer?

If neither of these blocks of licit users adopts digital currency, that leaves only criminals and tax evaders keen to use a new central bank digital currency. For a central banker who is advocating the stuff, that's not a very firm political leg to stand on. In sum, Carney has got it all wrong. A central bank digital currency is less likely to have a massively disruptive effect than it is to arrive stillborn.

PS: Thanks to Antti, Oliver and the rest on the discussion board for helping me think about this more concretely.

Wednesday, April 11, 2018

Moneyness = 22?

Courtesy of Kerry Taylor's twitter feed, here is a chart which was presented during a recent investing conference in Toronto. Apparently bitcoin has a moneyness score of 22 while cowry shells ring the bell at 15, both of them exceeding the moneyness of U.S. dollars at 13. The presentation that contains the chart was created by angel investor Sean Walsh and is available here.

Since my blog is called moneyness, and I've written quite a lot on this topic, I feel somewhat obligated to chime in. Let's start with the good bits about the chart. Instead of classifying items as money-or-not, we can appraise objects by their degree of moneyness. Because every valuable object or instrument is exchangeable, some more easily than others, everything lies somewhere on the money spectrum. The diagram below illustrates this idea. This way of looking at things can provide some insights that we don't normally get when taking the money-or-not approach, and its nice to see that folks like Walsh are using it. (For a longer explanation of moneyness, go here).

Now the not-so-good bits. Let's go and see what Walsh means by the term moneyness. On page 14 he lists six characteristics of money including scarcity, durability, divisibility, recognizability, fungibility, and tranportability. Walsh compiles an instrument's moneyness score by assigning a value from 0-4 for each characteristic and then summing this up. The maximum score is 24, with bitcoin losing just a point on durability and fungibility. He gives no explanation for how or why some instrument might get a 3 for, say, recognizability instead of a 4, so I guess we'll just have to assume he has a consistent method for rewarding points.

There are two reasons why I disagree with this approach. First, even if we accept Walsh's definition of moneyness and his choice of rankings for each instrument, his list of attributes is incomplete. It is missing one of the most important ones: price stability. When people accumulate balances in anticipation of spending needs, they expect those balances to hold their value for a few days, maybe weeks. If the medium's purchasing power is volatile, then there is a risk that the stuff in their wallets won't allow them to meet tomorrow's spending requirements, which means it isn't doing a very good job as a medium of exchange. Bitcoin probably has the lowest stability of the instruments in the chart. 

My second and more important criticism has to do with the way that Walsh measures moneyness. In a hard science like chemistry or geology, ranking each objects' physical characteristics might pass muster. For instance, geologists use the Mohs Hardness Test, a scale from 1-10 for testing the resistance of a mineral to being scratched. Walsh is running something like the Mohs Hardness Test, except for monetary instruments.

But economics involves humans. And in economics, we are not interested in the physical characteristics of the goods and services people buy, say how hard a mineral is, or how cushy a couch is, or how fast a car can go. Rather, we are interested in the subjective evaluation economic actors place on those objects and the manifestation of these preferences in the form of market prices.

So the way to accurately measure moneyness isn't to design the equivalent of Mohs Hardness Test for monetary instruments, but rather to find out what price people actually put on that moneyness. One way to do this is by asking how much compensation people would expect to earn if they were to give up an object's moneyness for a period of time. More specifically, say you are offered a deal to buy one bitcoin but are prohibited from selling that bitcoin for one year. How much less would you be willing to pay for this locked-in bitcoin than a regular bitcoin that you will probably hold for at least one year anyways? If a locked-in bitcoin is worth, say, $500 less to you than a regular bitcoin, that means that you place $500 on a regular bitcoin's one-year tradeability, or its moneyness.     

We can also think about moneyness in terms of interest rates. What rate would you need to earn on a locked-in bitcoin to compensate you for the nuisance of giving up its ability to be used as an exchange medium? 10%? 5%? The extra interest you expect on locked-in bitcoin is the degree to which you value a regular bitcoin's tradeability, or moneyness, over that time-frame.

The price of a dollar's moneyness is easy to measure. Someone who will have a spare $10,000 on hand for the next year can hold it in a government-insured chequing account and earn 0% or they can lock that amount into an insured term deposit and earn around 0.85% (I'm using Canadian numbers for non-cashable 1-year GICs). By locking in the $10,000, an individual's ability to mobilize these dollars as a medium for making payments has been effectively destroyed for 365 days. They cannot buy stocks or bonds with it, nor convert it into cash, nor purchase peaches, tables, labour, travel, etc. Their dollar are inflexible; they have no moneyness.

People are willing to accept this burden but only if they are compensated to the tune of 0.85%. Put differently, the 0.85% rate represents a large enough carrot that marginal depositors are roughly indifferent between holding money in a chequing account for a year or locking it in. So if $10,000 in a term deposit provides a pecuniary return of $85, then $10,000 dollars held in a 0%-yielding chequing account provides around $85 in non-pecuniary monetary services, or moneyness, over the course of the year.

We can also go through this process with gold. Head over to Kitco and you can see that the 12-month lease rate is at 0.2%. Say you are hoarding $10,000 in gold under your mattress. If you are willing to forfeit the ability to make any transactions with your $10,000 stash for one year, a bank will compensate you with $20 ($10,000 x 0.2%) for your pains. Put differently, $20 is the amount that the bank needs to provide the marginal gold hoarder to tempt them into giving up the moneyness of gold. (The implied moneyness of $20 is far less than the $85 a Canadian chequing account offers, contrary to Walsh's chart, which ranks gold above dollars. Note that I am ignoring storage costs.)

To carry out this measurement for bitcoin, we'd have to determine what sort of rates a large international bank provides to bitcoin term depositors. I doubt this measurement can be made since reputable banks don't deal in bitcoins. So bitcoin's moneyness is not 22. We have no real idea what it is.

Tuesday, March 27, 2018

More fiatsplainin': let's play fiat-or-not

The (Great) Tower of Babel, 1563, Bruegel the Elder. "Therefore is the name of it called Babel; because the Lord did there confound the language of all the earth"

People bandy the term fiat currency around a lot, but what exactly does it mean? None of us wants to live in a Babel where people use fiat to indicate twenty different thing. So let's try to zero in on what most people mean by playing a game called fiat-or-not. I will describe a monetary system as it evolves away from a pure commodity arrangement and you will tell me when it has slipped into being a fiat system. (The technique I am using in this post cribs from a classic Nick Rowe post).

So let's start the game.

1) An economy in which gold coins circulate as the medium of exchange.

Fiat or not? I think we can all agree that there is nothing fiat at all here. (For simplicity's sake let's assume for the duration of this post that taxes can be paid with anything, and that there is no legal tender.)

2) A government-owned central bank begins to issue banknotes that are redeemable into a fixed amount of gold. Owners of banknotes need only line up at the central bank's redemption window to convert their $1 notes into 1 gram of the yellow metal. The central bank ensures that its vaults contain 100% gold backing for its notes.

Fiat or not? Some people associate fiat with the invention of paper money or IOUs, but in general I don't think very many of us would say that these banknotes qualify as fiat.

3) The central bank sells off a chunk of its gold and invests in safe bearer bonds. Its banknotes are no longer 100% backed by gold coins, but are backed 70% bonds/30% gold. The central bank continues to redeem notes on demand with gold at a rate of $1 to 1 gram.

Say the public suddenly wants to hold more coins. A lineup develops at the central bank's redemption window and eventually the central bank uses up its coin reserves as it meets redemption requests. To continue meeting additional requests, it need only sell some of the low-risk bonds from its vault and use the proceeds to buy additional gold coins.  

Fiat or not? Since low-risk bonds have now become part of the backing for the banknote issue, a few readers may choose step 3 banknotes as the entry point for fiat money. But this would be unconventional, since most note-issuing central banks in the 1800s were running this sort of 70%/30% system, and we usually call the monetary system that prevailed in the 1800s a gold standard, not a fiat standard.

4) The central bank announces that it  will undergo extensive renovations. As a result, its redemption window will have to be shut for two months. People can no longer redeem their $1 for 1 gram of gold on demand, but will have to wait until the renovations are over.

Fiat or not? Two months is a long time. But it could be that the central bank already closes its doors on the weekends anyways, banknotes being inconvertible for 48-hours. I doubt many of us would describe the weekend as a fiat currency episode. Should we think of the renovation closure as an extended weekend, or is it long enough that it generates fiat money?

5) Unfortunately the central bank chose an incompetent construction company. Renovations will take another two years!

To make up for the inconvenience of the redemption window being closed for such a long time, the central bank promises to send agents to the local gold market who will ensure that the market rate stays fixed at $1/gram. These agents will buy & sell whatever amount of gold is necessary to maintain the peg (by selling and buying banknotes).

Fiat or not? Thanks to the strategy of buying and selling in the local gold market, the $1/gram price holds just as well as it did in steps 2 and 3. So the public notices no difference in the purchasing power of the money in their wallets. On the other hand, two years without a redemption window at the central bank may be long enough for many readers to tick the fiat money box.    

6) The central bank is still undergoing renovations, but instead of dispatching agents to the market to buy and sell gold to enforce the peg, they go with bonds in hand.

If the market price for gold threatens to rise from $1/gram to $1.01/gram, because there is too much money chasing too few goods, the agents sell bonds and withdraw banknotes, thus reducing pressure on the exchange rate and bringing it back to $1/gram. And when the exchange rate threatens to fall below $1/gram to $0.99/gram, because there is too little money chasing goods, agents buy bonds with banknotes.

Fiat or not? Not only are notes not redeemable in gold, but now the central bank no longer operates directly in the gold market. With this step we are getting a bit closer to modern central bank money. The Federal Reserve, the Bank of Canada, and other major central banks all regulate the purchasing power of money by purchases and sales of bonds. The $1/gram peg still holds thanks to bond purchases and sales, so step 6 money does almost everything that step 2 and 3 money does.

7) With the renovation dragging on, the central bank decides that it doesn't need a redemption window after all. So what was initially a temporary suspension of convertibility becomes permanent. But the central bank continues to send agents to the market to buy or sell whatever quantity of bonds are necessary to maintain the $1/gram peg.

Fiat or not? You tell me. Perhaps permanent inconvertibility is the very definition of fiat. However, if steps 2-6 didn't qualify as fiat money, because gold stayed at $1/gram, why would step 7 be any different?

8) The central bank decides that, rather than fixing the market price of gold at $1/gram, it will set the market price of a typical consumer basket of goods and services (i.e. meat, car repairs, school, etc). 

This is a bit trickier to think about than the other steps. So for example, say that the central bank is currently setting the price of gold at $1/gram. And people can buy a consumer basket for $1000. But the price of that basket starts to rise to $1010, $1020, and then $1030. To stop this inflation, the central bank will announce its intention to reduce the price of gold to $0.99/gram. It does this by selling bonds and withdrawing money from the system, so that there is less money chasing goods. It keeps repeating gold price decreases/money withdrawals until it has successfully reigned in the inflation and brought the consumer price basket back to $1000. The net effect is that consumers are always guaranteed that the money in their pocket has constant purchasing powe

Fiat or not? This is pretty much the monetary system we have now in the U.S. and Canada where central banks target inflation. Well, there are a few small differences. Instead of temporarily setting the price of gold in order to regulate the value of a consumer price basket, the Fed and Bank of Canada temporarily set the price of a very short-term debt instrument to hit their target for the basket. And rather than shooting for constant consumer goods and services prices, these central banks prefer one that shrinks by 2% a year.

Given that step 8 describes something close to modern money, and it is common practice to refer to modern money as fiat, then it would only make sense that many readers raise their hands at this point. Complicating matters is that step 8 money isn't really that different from steps 2 to 7. After all, the central bank is establishing a fixed price for banknotes, the only difference being that the fix has been adjusted from gold to a basket of consumer goods and services. 

9) The central bank donates all of its assets to charity, closes its doors and shuts down for good. But it leaves all its banknotes outstanding. Money floats around the economy without a tether to reality. Or as Stephen Williamson says, money is a bubble.

Fiat or not? By this stage, everyone will probably have ticked the fiat money box. 


Here is a collection of unconnected thoughts on the fiat-or-not game.

A) My guess it that readers will have chosen different stages as their preferred debut for fiat money. This is a bit tragic, since with no commonly-accepted definition for the term, most debates about fiat money have been and will continue to be meaningless.

B) We apply our definitions like cookie cutters to the real world. So if you chose step 7 (when banknotes became permanently irredeemable) as your flipping point, then 1971 would be a very important date in your scheme of the world since this is when the U.S. permanently removed gold convertibility.

But if you chose step 9 as your transition point to fiat, then the global monetary system is not currently on a fiat standard, since central banks have neither closed their doors nor donated their assets to charity. So 1971 really isn't an interesting date. I'm aware of only one country on a step 9 fiat standard: Somalia. Its central bank burned down yet Somali shilling banknotes continued to circulate. And ironically enough, if we choose to adopt a step 9 definition of fiat money, then bitcoin—which was designed to destroy central bank "fiat" money—is itself fiat, because it is unbacked, whereas most central bank money is not fiat.

What I've described is the Borges problem. Categories pre-digest the world for us. We get very different results depending on what definition we use and how we apply it to the world.

C) I think many readers associate fiat with hyperinflatable. For instance, here is Dror Golberg:

Readers who conflate fiat and hyperinflatable will probably have played the fiat-or-not game by gauging each step to see if it introduced (or removed) a set of features perceived to be conducive (inhibitory) to high inflation. They probably toggled the fiat button somewhere in the murk of temporary inconvertibility (step 4) and permanent inconvertibility (step 7). The thinking here is that convertibility into specie imposes a more imposing restriction on a central bank than a mere promise to hold gold's value at $1/gram by using open market operations (step 6). With the removal of convertibility, hyperinflatability is activated and thus money has become fiat.

There are certainly some good historical reasons for assuming that inconvertibility leads to hyperinflatability. Some of the most famous hyperinflations occurred after redemption was removed, including John Law's paper money scheme, the American Greenback episode, and the Wiemar inflation. But there is no inherent reason that these systems must lead to hyperinflation, or that step 1 (coin-based systems) and step 2 (fully convertible) systems aren't themselves hyperinflatable. In the case of coin-based systems, all that it takes is a rapid series of reductions in the silver content of coins to set off inflation, Henry VIII's consistent debasement of the English coinage being one example. And there is no reason that a fully convertible step 2 banknote system can't undergo a series of large devaluations leading to hyperinflation. 

D) Fiatness, fiatish? If we can't agree on what constitutes fiat-or-not, maybe we can agree that there might be a fiat scale, from pure fiat to not fiat at all, with most monetary systems existing somewhere in between. I am already on record advocating moneyness over money, so this fits with the general them of the blog. On the other hand, fiatness seems a bit of a cop-out.

E) We don't need gobbledygook like fiat. The term carries too much baggage. Let's select a more precise set of words, then apply them to the real world in order to understand what our monetary systems were like, how they are now, and where we are going. Until we settle on these words, let's avoid all conversations with the term fiat in them.

P.S. I have a recent post about the desirability of coin debasements at the Sound Money Project and another post on money as a measuring stick at Bullionstar. 

Wednesday, March 21, 2018


I am a big fan of coinsplainers like Andreas Antonopoulos. Listening to Andreas explain how bitcoin works is a great learning opportunity for folks like myself who know far less about the topic. I am less impressed when bitcoiners engage in fiatsplainin', since they generally have an iffy understanding of the actual financial system and central banking in particular.

So for the benefit of not only bitcoiners, but anyone interested in the topic of money, I'm going to fiatsplain' a bit. (I really like this term, I got it from an Elaine Ou blog post)

Paul Krugman recently had this to say about the difference between bitcoin and fiat money:
"So are Bitcoins a superior alternative to $100 bills, allowing you to make secret transactions without lugging around suitcases full of cash? Not really, because they lack one crucial feature: a tether to reality.
Although the modern dollar is a “fiat” currency, not backed by any other asset, like gold, its value is ultimately backed by the fact that the U.S. government will accept it, in fact demands it, in payment for taxes. Its purchasing power is also stabilized by the Federal Reserve, which will reduce the outstanding supply of dollars if inflation runs too high, increase that supply to prevent deflation.
Bitcoin, by contrast, has no intrinsic value at all. Combine that lack of a tether to reality with the very limited extent to which Bitcoin is used for anything, and you have an asset whose price is almost purely speculative, and hence incredibly volatile."
Now if you've been reading my blog for a while, you'll know that I agree with Krugman's point that bitcoin lacks a tether to reality while a banknote doesn't. He mentions two forces that anchor a $100 banknote, or provide it with intrinsic value: tax acceptability and a central bank's guarantee to regulate its quantity. Let's explore each of these anchors separately, starting with tax acceptability.


The idea that taxes can determine the value of a fiat currency is easier to grasp by looking at currencies issued during the American colonial era. Coins tended to be scarce in the 1700s and there were few private banks, so the legislatures of the colonies issued paper money to meet the public's demand for a circulating medium. They had a neat trick for ensuring that this paper money wasn't deemed worthless by citizens. A fixed quantity of paper money was issued concurrently with tax legislation that scheduled a series of future levies large enough to withdraw each of the notes that the legislature had issued. This combination of a fixed quantity of notes and future taxes of the same size was sufficient to give paper money value, since the public would need every bit of paper to satisfy their tax obligations.

Examples of colonial currency (it's worth enlarging this image to see the detail) From: Early Paper Money of America

Crucially, once a colonial government had received a note in payment of taxes, it removed said note from circulation and destroyed it. If the government re-spent notes that had already been used to discharge taxes, this would be problematic. The tax obligation would be more-than-used-up, leaving no reason for the public to demand outstanding banknotes. Krugman's "tether to reality" would have been removed.


The modern day version of Krugman's tax acceptability argument is a bit more complicated. For starters, no one actually pays their taxes with banknotes. Rather, the tax acceptability argument applies to a second instrument issued by central banks otherwise known as reserves (in the U.S.) or settlement balances (in Canada). All commercial banks keep accounts at the central bank, these accounts allowing them to make instant electronic payments to other banks during the course of the business day, or to the government, which typically will also have an account at the central bank.

When Joe or Jane Public are ready to settle their taxes, they initiate a set of financial transactions that ultimately results in their bank depositing funds on their behalf into the government's account at the central bank. To satisfy the public's demand to make tax payment, commercial banks will want to have some central bank settlement balances on hand. So the existence of taxes "drives" banks to hold a certain quantity of central bank settlement balances, thus generating a positive price for these instruments. And since a banknote is in turn tethered to a central bank deposit via the central bank's promise to convert between the two at par, by transitivity the banknote is also tethered.

Unlike the colonial era, however, the tax authority—the government—can't destroy money. The government can either accumulate central bank deposits, or spend them, but it can't cancel them. What generally happens with the government's account at the central bank is that as soon as it is topped up with some tax receipts, they get quickly spent on government programs, salaries, and other expenses. So these funds simply boomerang right back into the accounts that commercial banks keep at the central bank, undoing the tethering that is achieved by tax acceptability.

Put differently, for every bank that demands settlement balances to pay taxes, and thus help gives those balances value, there is a government official who spends them away, and negates this value. So government taxes by themselves don't anchor modern central bank money.

To really anchor the value of central bank money, the government needs to withhold from spending the money it has received from taxes. The more it resists spending incoming tax flows, the more balances accumulate in its account at the central bank. If the government keeps doing this, at some point almost every single deposit that the central bank has ever issued will have been sucked up into the government's account. With almost no deposits remaining for paying taxes—and thus no way for the public to avoid arrest for failure to meet their tax obligations—the value that banks collectively place on deposits will reach incredible heights.

And that explains how tax acceptability (combined with a strategy of not spending taxes received) can provide modern fiat money with backing sufficient to generate a positive price.


Let's turn now to Krugman's second reason for central bank money having intrinsic value, the central bank itself. As I said earlier, a government can freeze deposits by accumulating them, but it can't destroy them. The only entity that can destroy money is the central bank. It achieves this is by conducting open market sales of bonds and other assets. When it sells a bond to a bank, the central bank gets one of its own deposits in return, which it proceeds to destroy.

Imagine that banks collectively decide they have too many central bank deposits and start to sell them (a scenario I discussed here). This sudden urge to rid themselves of money will cause inflation. In a worst case scenario, they will get so desperate that the purchasing power of money falls to zero. The central bank can counter this by selling assets and destroying deposits. In the extreme, it can sell each and every one of the assets it owns, shrinking the deposit base to zero. Its actions will drive the value of deposits into the stratosphere, since banks need a token amount to make interbank payments.

And that, in short, explains how central banks can provide dollars with backing sufficient to generate a positive price.


Which of Krugman's two forces—tax acceptability or a central bank's guarantee to regulate the quantity of money—is more important for imbuing little electronic bits with value?

We know that a government can anchor a fiat money purely through tax acceptability. Colonial money proves it. (Here is another example from the Greenback era) But can a fiat currency be anchored solely through the actions of the central bank, without the help of tax acceptability? Let's set the scene. Imagine that the government has unplugged itself from the central bank by closing its account and instead opening accounts at each of the nation's commercial banks. Since all incoming tax receipts and outgoing government payments are now made using private bank deposits, the government no longer generates a demand for central bank settlement balances.

This "unplugging" needn't drive the value of central bank money to zero. The central bank has assets in its vault, after all, so any decline in the value of central bank money can be easily offset by an appropriate set of central bank open market sales and concomitant reductions in the quantity of deposits. So the answer to my question in the previous paragraph is that money doesn't require tax acceptability to have intrinsic value. Tax acceptability is sufficient, but not necessary.

That being said, on a day-to-day basis the value of modern central bank money is regulated by a messy combination of both factors. Money is constantly flowing in and out of the government's account at the central bank, and this can have an effect on the purchasing power of money. Likewise, central bank open market operations are frequently conducted on a daily basis in order to ensure the system has neither a deficiency nor an excess of balances. It's complicated.

And that ends this episode of fiatsplainin.' Fiat money is indeed backed and has intrinsic value, as Krugman says, and it does so for several reasons.

PS. If you are interested in colonial currency, you should read some of Farley Grubb's papers.


On Twitter, someone had this to say about my post:
฿ryce gives me the perfect opportunity to keep fiatsplainin'. Contrary to ฿ryce's claim, the fact that Arizona plans to accept tax payments in the form of bitcoin does not provide bitcoin with a tether to reality. For every bitcoin that Arizona accepts, it will just as quickly spend it away. The first is undone by the other. You'll notice that this is the same reason I gave for modern central bank money not necessarily being anchored by tax acceptability; whereas taxes vacuum up central bank money, government officials typically reverse this vacuum by quickly spending it, so the net effect is a wash.

To tether central bank money to reality, governments need to not only make it tax acceptable but also  be ready to let those balances pool up in its account, thus setting a limit on the overall supply of balances. Likewise with bitcoin. If the Arizona government were to accumulate incoming bitcoins as part of an overall policy of never spending them, then it would be removing bitcoins from circulation, in essence "destroying" them. And this would provide bitcoin with a true anchor. Of course the Arizona government isn't going to do this. It will want to rid themselves of bitcoins the moment it gets them.

Wednesday, March 7, 2018

Indians' "ill-informed notions" concerning the legitimacy of ₹10 coins

The BBC has an interesting story about India's coinage. Apparently more and more Indians  believe that the ₹10 coin is not real, or that it has been banned by the authorities, and as a result they are unwilling to accept them in trade. Doubts about the ₹10 coin have been emerging for several years now: Amol Agarwal has covered the story here, here, and here.

This is an excerpt from the BBC article:
"Nobody accepts the coins - grocery shops, tea stalls, nobody accepts it", an auto rickshaw driver in the southern state of Tamil Nadu told BBC Tamil.
In the southern city of Hyderabad, a young girl told BBC Telugu she had been saving up to buy her brother a gift but several shop owners wouldn't take her 10 rupee coins.
A man on his way to a job interview was forced to get off the bus because the conductor wouldn't accept 10 rupee coins, the only currency he had.
"They say it's because the other passengers don't accept the coins in return", explains a shop owner who also said bus conductors wouldn't take the coins.
The Reserve Bank of India (RBI) has twice addressed the public's worries about the ₹10 coin. In a 2016 announcement it begged Indians to ignore "ill-informed notions" concerning the legitimacy of ₹10 coins and to continue to "accept these coins as legal tender in all their transactions without any hesitation." More recently, in a January notice, we learn that the RBI has issued fourteen different designs for the ₹10, all of which are "legal tender and can be accepted for transactions."

What are the underlying reasons for Indians' fears? One interesting fact about the ₹10 coin is that it is relatively new, having been introduced back in 2009. People are always skeptical about new monetary instruments, which generally take a long time to acquire trust.

Another interesting fact is that in addition to minting a ₹10 coin, the RBI also prints a ₹10 banknote. The ₹10 banknote has a long history, having debuted before independence in 1947. Below is a chart showing how many of each instrument is in circulation. The year-over-year net increase in banknotes continues to outpace the increase in coins by a large amount, indicating that  Indian's have a preference for the paper version of the ₹10.

I think there is an easy explanation for the ₹10 coin's loss of currency. Because the ₹10 coin and ₹10 note are perfect substitutes, and converting between them incurs no conversion costs, there is no disciplining mechanism to prevent irrational worries about the newer of these two instruments from crippling its usage. Put differently, hating new ₹10 coins doesn't impose any costs on the hater as long as an equivalent banknote can be used. If there was no such thing as the ₹10 banknote, then anyone who refused to use the ₹10 coin would face much higher costs for being unreasonable. After all, holding two ₹5 coins or five ₹2 coins in the place of a ₹10 coin is inconvenient.

The denomination at which a monetary system switches from coins to notes is referred to by Rocheteau and Lotz (pdf) as the coin-note frontier. In Canada, for instance, the frontier lies between the $2 coin and $5 note, while in Switzerland it lies between the 5 franc coin and 10 franc note. Most frontiers (like Canada's and Switzerland's) are staggered—the largest coin is smaller than the smallest note. This staggering makes a lot of sense. Why should both the nation's mint and its printing presses incur the fixed costs of producing the same unit when one will suffice? Consider too the waste incurred in the doubling-up of the tasks of distributing, sorting and handling a coin and note of the same denomination.

Unlike most countries, India has an even coin-note frontier. For some reason, the Indian monetary authorities have decided to have both the mints and the presses replicate the same task of producing the ₹10. Interestingly, India isn't alone. The U.S.'s largest coin is $1 while the smallest note is $1.

The US's $1 coin, introduced in 1979 and referred to as the Susan B. Anthony dollar, is commonly considered to be a major monetary failure. I wrote about it here. $1 coins have proven to be unpopular with the American public, huge amounts of them accumulating in vaults at various Federal Reserve banks. Because the US monetary authorities decided to introduce the $1 coin without removing the $1 bill, the public was given a choice between a perceived "good" currency, the existing and comfortable note, and a "bad" currency, an unfamiliar coin. They took the less costly route and stuck with the "good" notes. My guess is that the very same forces that doomed the $1 coin could end up killing off the ₹10 coin.

The failure of the $1 and ₹10 coins is unfortunate. As Rocheteau and Lotz point out, replacing low denomination notes with coins is a good idea because the the cost of keeping bills in circulation is greater than the cost of servicing coins. While coins are more expensive to produce, they last much longer than bills.

So not only are the US and India doubling up their costs by having both the mint and printing presses produce the same instrument, but at the same time the decision to keep the note in circulation means that the more efficient instrument—the coin—is destined to fail. The Reserve Bank of India blames the public's "ill-informed notions" for the ₹10 coin's loss of currency. But perhaps it should be blaming itself for providing the right conditions that allow for the spread of these ill-informed notions. Remove the ₹10 note and the problem will be fixed.

Amol Agarwal has some comments here.

Friday, March 2, 2018

The odd relationship between gangster and central banker

In my recent post for the Sound Money Project, I touched on the odd relationship between central banker and gangster. I want to focus a bit more on this relationship.

An awkward truth of central banking is that one of the central bank's most important lines of business—the business of providing cash, specifically high denomination banknotes—primarily serves hoodlums, gangsters, tax evaders, and the mafia. Yes, non-criminals certainly make some use of high denomination banknotes, say a few notes hidden in the cookie jar in case the electricity goes down. But the largest base of users is comprised of folks who hold notes—not in cookie jars—but by the suitcase full; criminals. Banknotes are anonymous after all, so they are an excellent way for criminal organizations to make large-scale transactions without being traced.

Providing criminals with high-denomination banknotes is a lucrative line of business. For each $100 note put into circulation, a central bank holds $100 worth of interest earning assets in its vaults. Since note holders don't have the right to receive any interest, the central banks gets to keep all this interest income for itself.

For instance, by the end of 2016 the Bank of Canada had placed $80.5 billion worth of banknotes into circulation. Large denomination banknotes—the $50, $100 and $1000 notes—accounted for $58.4 billion of this, or around 72% of all banknotes. The assets standing behind all outstanding banknotes allowed the Bank of Canada to earn $1.53 billion in interest in 2016. Of this amount, around $1.1 billion (72% of $1.53 billion) can be attributed to high denomination banknotes, the majority of which comes courtesy of the largest holders of high denomination notes: gangsters.

So you can begin to see why the Bank of Canada might not want to get out of the business of producing $50, $100, and $1000 notes. $1.1 billion is a lot of profit! Of course, were the Bank to get out of producing high denomination notes altogether, it wouldn't forgo the entire $1.1 billion in yearly income. Criminals might choose to use $10 and $20 notes in the place of the demonetized high denomination notes. However, $10s and $20s are a bulky way to store value. They surely wouldn't be capable of recapturing all of the criminal wealth formerly held in the form of $50, $100, and $1000 notes. Which means that the total amount of banknotes outstanding would fall and Bank of Canada profits would shrink.

Why might central bankers care about their profits? As I wrote here, any government bureaucrat who can provide their master with an ongoing revenue stream will always have more say in their department's fate than a bureaucrat who has to ask for funding each year. And of all government bureaucrats, none is more jealous of their independence than the central banker. The process of ratcheting the interest rate lever higher or lower requires a complete absence of political meddling, so say central bankers. One might imagine that this autonomy is worth so much to central bankers that it justifies taking on a clientele dominated by gangsters.

There is a better reason for why it might be in the public interest for central bankers to continue serving criminals with high denomination banknotes. Consider the fact that if high denomination notes were to be rescinded, criminals would simply use other forms of payment in their place. If the substitute payments medium that criminals select places a new and extremely onerous set of burdens on society, then maybe the public provision of high denomination notes should not be discontinued.

What alternative payments media might criminals use in the place of $100 and $50 notes? In his screed against high denomination banknotes, Ken Rogoff suggests that gold, uncut diamonds, and bitcoin might become popular as a criminal payments media. The fact that these instruments are cumbersome relative to cash would make criminals easier to catch, and Rogoff claims that the crime rate might even drop.

In a provocative article, James McAndrews counters that rather than turning to commodities, criminals will instead select private debts as their preferred payments medium. A thief who sells stolen goods to a fence would accept some sort of IOU as payment rather than cash or diamonds. This IOU wouldn't be anonymous. Like any debt, the debtor and creditor would be a matter of record. But as long as the system of debts is secret—i.e. only criminal participants can see the record—then the users can't be tracked by the authorities, like cash.

When an IOU defaults, the traditional legal system provides a means for sorting things out. But this system would be out of bounds to criminals trafficking in IOUs. What is required is some sort of underground administrator or third-party to act as arbiter. According to McAndrews, the party that is likely to emerge as enforcer of criminal debts is organized crime: the mafia. 

In addition to enforcing IOUs, the mafia would also be in a position to fabricate new IOUs for use in the criminal monetary system. McAndrews uses the example of inflated invoices. The mafia would coerce legitimate businesses into writing IOUs, or invoices, for goods they never bought, or bought  at inflated prices. These invoices would circulate among criminals as money. To assure that the police ignored their extortion of legitimate business, the mafia would resort to stepped up bribery of the police.

All of this changes the calculus of a central bank withdrawal from the business of providing criminals with banknotes. Sure, a demonetization of high denomination notes might lead some gangsters to go legit because the lack of $100 and $50 notes makes their business too expensive to operate. But a whole new range of crimes could emerge. Violence could grow as the mafia executes defaulters in order to maintain the sanctity of the new IOU payments system that has taken the place of high denomination banknotes. Legitimate businesses could get blackmailed into feeding the criminal monetary system, those run by immigrants likely being the most vulnerable. And police departments will be corrupted.

McAndrews uses the public provision of free condoms and clean needles as an analogy. Restrict free condoms and it is possible that the rate of sexual intercourse goes down. But surely there will be an increase in unsafe sex, unplanned pregnancy, and sexually transmitted diseases. As for the provision of clean needles, restrict it and heroin use might fall. However, the prevalence of HIV will rise. Both unsafe sex and dirty needle usage impose costs not only on those directly afflicted but also indirectly on us—i.e. taxpayers who pay increased health care expenditures.

Likewise with cash. A restriction of $50 and $100 notes could very well lead to attrition in the ranks of existing criminals, as Ken Rogoff reasons. However, this could be twinned with an increase in mafia activity and the potential subordination of us—i.e. legitimate business—to the needs of the underground payments system. Keeping high value banknotes may thus be the wise decision, in the same way that choosing to keep free condom and clean needle programs going makes everyone's lives better off.

Tuesday, February 20, 2018

Cash, cat, and mouse

Bruno Liljefors, 1899, link

The tax authority and the tax payer are engaged in an age-old cat and mouse game, tax payers trying to perfect tricks that allow them to pay as little tax as possible and the tax authority trying to close these loopholes. Retail cash payments are one of the fields on which this battle is waged. It's interesting to see how sophisticated this cat and mouse game has become.

There are two weak points in the sales process that allow cash-accepting retailers to avoid paying sales taxes or VAT. The first weak point is at the very outset of a payment. When a customer pays with cash, the person behind the till can avoid ringing up the payment. Without a record of the payment having been made, the retailer needn't pay tax.

But even if a retailer rings up all cash payments and provides receipts, they can still avoid paying taxes. At the end of the business day, they need only doctor the cash register's data using a zapper—add on hardware or software designed for this purpose—in effect purging all or a portion of the cash payments registered during the course of business. With the only record of that day's cash payments now being the paper receipts held in customers' wallets—most of which will have been thrown away—the retailer needn't worry about the tax authority discovering the doctoring. (Erasing card based payments is much riskier for the retailer because a paper trail still exists with the card-issuer.)

Tax authorities have been targeting the second point of weakness for a few decades now by requiring retailers to use certified cash registers that have tamper-proof memory units. These are variously known as a fiscal control units, electronic tax registers, or fiscal tills. These tills are designed in such a way that any attempt on the part of the retailer to break into its memory using a zapper or some other technique will be discovered. Additionally, these units have the potential to be connected directly to the tax authority, allowing for instantaneous transmission of sales data and constant real-time tax auditing. That sounds a bit intrusive, no?

Below is a chart from the IMF showing nations that have implemented fiscal till plans:

Source: IMF, Electronic Fiscal Devices, 2015

Progressing to the next stage of the cat and mouse game, retailers will try to evade tamper proof memory units in the cash register by making recourse to the first weak point in the sales process; not entering the transaction into the cash register in the first place. Recognizing this, the tax authorities who have implemented fiscal till schemes have made it illegal to not issue a sales receipt. But illegality doesn't seem to me like a big hindrance to a retailer who has already set their mind on evading taxes.

One neat trick to get retailers to provide receipts—and therefore run all transactions through the tamper-proof cash register—is to recruit the customer into the cat and mouse game as helper. Public information campaigns exhorting people to ask for receipts are one technique. But the more interesting trick is implementing a tax-receipt lottery. All invoices issued from the tamper-proof cash register come with a unique lottery number. Anyone who keeps their invoices will be able to participate in a periodic lottery. Customers thus have an incentive to ask the retailer for a receipt, obliging the retailer to run the transaction through the fiscal till.

Taiwan implemented the first tax receipt lottery back in the 1950s, the Uniform Invoice lottery. I've included a picture below, and here is the website. In the last fifteen years, a number of nations have begun to copy it including Czech, Slovakia, Slovenia, Malta, Portugal, Poland, China, Sao Paulo, and Lithuania.

Taiwan sales receipts with lottery numbers on them

The next stage of the cat and mouse game occurs as the retailer, desperate to adapt to the government's crafty invoice lottery, tries to coax the customer over to his side. On a $50 meal, a restaurant may be able to save $2.50 in tax (assuming a 5% tax rate) if the the fiscal till is avoided. If the restaurateur says that he will share some of this savings with the customer, he may be able to induce her to not ask for an invoice and thus avoid the till. The amount of money he must dangle in front of her will have be large enough to compensate her for the foregone fun of playing the lottery, potential lottery winnings, and guilt.

China is the most interesting example of the cat and mouse game being played at this level. To incentivize cash-paying customers to ask for invoices, or fapiao, the Chinese authorities have created a scratch and win game. Restaurateurs have reacted by offering customers a free soda, or a discount, if they don't ask for the fapiao. Presumably the value of a soda is just sufficient to compensate the customer for foregoing the lottery. More entertaining accounts of fapiao here and here

I'm sure these methods of attacking tax avoidance work to an extent. In Québec, for instance, as of March 2016 the tax authorities say that they have recovered CAD$1.2 billion in taxes following the introduction of fiscal tills in the restaurant industry. However, I'll hazard that the biggest determinant of tax avoidance is good government. If people trust the government to do smart things with tax revenues and they don't see evidence of corruption, then they will be more likely to view paying taxes and reporting on cheaters as one of their public duties.

Other sources:
OECD: Technology Tools to Tackle Tax Evasion and Tax Fraud (link)
Ainsworth: Québec’s Sales Recording Module (SRM) - Fighting the Zapper, Phantomware, and Tax Fraud with Technology (link)
Steenbergen: Reaping the benefits of Electronic Billing Machines (link)

Tuesday, January 30, 2018

The big ol' €500

Production of the European Central Bank's €500 notes is scheduled to come to an end later this year. But a chart of the quantity of €500 banknotes in circulation (see below) reveals something odd. The supply of €500s began to plummet way back in early 2016, long before note production was supposed to be halted. What gives?

It was back on May 4, 2016 that the ECB officially announced that it would stop printing and issuing the €500 note, one of the world's most valuable banknotes ranked by purchasing power. The reason it gave was concerns that the €500 "could facilitate illicit activities." You may remember that this was in the midst of ex-banker Peter Sands screed against high denomination notes, echoed by economist Larry Summers and later amplified by Ken Rogoff's book The Curse of Cash.

While the €500 is undoubtedly popular with organized crime, there is some evidence that regular people use €500s, as Larry White points out here. In the recently published survey on the use of cash by households in the euro area, 19% of respondents reported having a €200 or €500 in their possession in the previous year. A quarter of respondents held banknotes (they don't specify the denomination) as a precautionary reserve, with 12% of these reporting a stash greater than €1000. So that means that around 3% of Europeans keep a large hoard of notes under their mattresses. This presumably gives the €500 a role to play as a store of value. After all, hiding thirty €500s under the bed is more convenient than three-hundred €50s.

But concerns over illicit usage of the €500 won out. Issuance of new €500s is set to stop near the end of 2018, although after that date people will be free to continue holding existing €500s as a store of value or to buy things. Any note deposited in the banking system after that point will be sent to the ECB to be destroyed. With no new supply and a steady removal of existing €500 notes, the quantity outstanding after 2018 will steadily shrink.

Below, I've charted out the total value of euro high denomination banknotes in circulation.

Although the €500 has eight or nine months left before this deadline is reached, the supply has already fallen by around €50 billion from its peak level of €300 billion outstanding in January 2016. Has the ECB jumped the gun and already kiboshed the €500 without telling anyone?

Luckily, the ECB provides incredibly fine-grained data on banknotes. Not only can we get the total value of banknotes in circulation, but also the monthly flow of banknotes issued by the ECB to private banks and returned by private banks. I've charted these flows below.

No, the ECB has not jumped the gun. It continues to issue several billion euros worth of €500s each month (the black line). But whereas issuance tended to exceed note returns in the past—the result being growth in the total stock of €500s in circulation—the tables have turned and note returns (the grey line) have generally exceeded issuance since early 2016, and thus the stock has dwindled. So the observed decline in the supply of €500s is entirely the result of the public's preference to have less of them.

This highlights an important point that I often mention on this blog. One of the most popular motifs of central banks is that they print cash willy nilly, forcing it onto an unsuspecting and virginal economy. This wildly misses the mark. Central banks do not push banknotes into the economy. Rather, the public pulls banknotes out of the central bank into the economy and pushes them back to the central bank. Each month Europeans return whatever quantity of €500s they don't want to the banking system, commercial banks in turn forwarding this currency to the ECB. Others withdraw whatever amounts of €500s they desire from their bank accounts, private banks in turn calling on the ECB to provide sufficient €500s. The net effect is an increase or decrease in the total stock of €500 banknotes in circulation. The ECB itself has no direct control over the public's decision to build or diminish the total supply of €500s.

I suspect that the relatively large increase in €500 note returns since 2016 is due to worries of an aggressive demonetization. As the second chart shows, returns of €500s began to accelerate in February and March 2016, well before the May 2016 announcement date. At the time, hints of the €500's imminent demise were being leaked to the press. Now, imagine that you are the head accountant at a large criminal organization with multiple suitcases full of €500s. You are hearing rumours that something is about to be done to the €500 note. The worst case scenario is that the note is to be suddenly cancelled—or demonetized—by the ECB, the period for converting €500s into €100s and €200s limited to a harrying few weeks. If the conversion window is being monitored by the authorities, your organization's attempts to convert €500s into smaller denominations might be flagged for further inspection.

Given this scenario, you'd want to change your suitcases full of €500s into €100 and €200s as fast as possible, before the actual announcement hits. Otherwise your organization might end up forfeiting a large chunk of the value of those notes—and you might be fired, literally. So my guess is that the rumours surrounding the fate of the €500 probably caused a mini "banknote run" in the months prior to the May announcement. Even after the ECB assuaged worries about an aggressive demonetization by promising to exchange €500s for an unlimited period of time, note returns have remained high relative to issuance. This suggests that the underground market still has worries about a potential aggressive demonetization, and are shifting into safer alternatives.

Once the ECB stops issuing €500s at the end of this year, the pull-push mechanism I described above will cease to function. There are two ways to set monetary policy. The first way—the one that regulates all banknotes including the €500—is to fix the price and let the quantity fluctuate as the public pulls what it needs and pushes back what it doesn't. The other policy is to fix the quantity and let the price fluctuate. This is the policy governing assets like gold, or the S&P 500, or bitcoin. 

After 2018 the ECB will have switched from fixing the price of €500s to fixing their quantity. At that point, the price will become a floating one determined by public demand, just like gold or bitcoin or the S&P 500. The higher the public's demand for €500s, the more its price will rise relative to pegged banknotes like the €100. A few years from now, it might take six or seven €100s to buy one €500.

Thursday, January 25, 2018

Paying interest on cash

Freigeld, or stamp scrip, is designed to pay negative interest, but it can be re-purposed to pay positive interest.

Remember when global interest rates were plunging to zero and all everyone wanted to talk about was how to set a negative interest rate on cash? Now that interest rates around the world are rising again, here's that same idea in reverse: what about finally paying positive interest rates on cash? I'm going to explore three ways of doing this. As for why we'd want to pay interest on cash, I'll leave that question till the end.


The first way to pay interest on cash is to use stamping. Each Friday, the owner of a bill—say a $50 note—can bring it in to a bank to be officially stamped. The stamp represents an interest payment due to the owner. When the owner is ready to collect his interest, he deposits the note at the bank. For example, say that 52 weeks have passed and 52 stamps are present on the $50 note. If the interest rate on cash is 5%, then the banknote owner is due to receive $2.50 in interest.

Alternatively the note owner can collect the interest by spending the $50 note, say at a local grocery store. The checkout clerk will count the number of stamps, or interest due, and tack that on to the face value of the note. With 52 stamps, the owner of a $50 note should be able to buy $52.50 worth of groceries, not $50. After all, the store has the right to bring the $50 note to its bank and collect the $2.50 in interest for itself.

Stamped currency seems like a pretty big hassle to me. The clerk behind the counter must count out the stamps on the note by hand, and the owner of the note has to trek back and forth to the bank each week to get the stamp affixed. Instead, imagine that each banknote has a magnetic strip that records how long the bill had been in circulation. This would remove some of these hassles. Weekly trips to the bank for stamping would no longer be necessary, and a note reader installed at a bank or retailer would automatically record how much interest was due, precluding painstaking counting of stamps.

"They use this magnetic strip to track you." says Byers to Agent Scully, The X-Files

Apart from stoking conspiracy theories, there's still a major problem with a magnetic strip scheme. Because each note has entered circulation at a different time, each is entitled to a varying amounts of interest. And this means that banknotes are no longer fungible. Fungibility—the ability to cleanly interchange all members of a population—is one of the features of money that makes it so easy to use. Remove it and money becomes complicated, each piece requiring a unique and costly effort to ascertain its value.


Our second way of paying interest on money doesn't destroy the fungibility of banknotes. The central bank needs to sever the traditional 1:1 peg between deposit money and cash, and then have cash slowly appreciate in value relative to deposits.

For instance, a central bank might start by setting an exchange rate of $1 note = $1 deposit on January 1, but on January 2 it adjusts this rate so $1 note is equal to $1.0001 deposits, and on January 3 adjust this rate to $1:$1.0002, etc. So the cash in your wallet is benefiting from capital gains. By December 31, the exchange rate will be around $1 note to $1.0365. Anyone who has held a banknote for the full year can deposit it and will have earned 3.65 cents in interest, or 3.65%. 

The major drawback with this scheme is the calculational burden imposed on the population by breaking the convenient 1:1 peg between cash and deposits. Assuming that retailers price their wares in terms of deposits, anyone who wants to pay in cash will have to make a currency conversion using that day's exchange rate. For instance, if the central bank's peg is currently being set at $1 note = $1.50 in deposits, then a popsicle that is priced at $1 will require—hmmm... let me check my calculator—$0.667 in cash. Phones will make this exchange rate calculation easy, but it is still likely to be a bit of a nuisance.

There are other hassles too. Would a capital gains tax have to be paid on the appreciation of one's cash? How would existing long-term contracts deal with the divergence? For instance, if my employer is paying me $50,000 per year, obviously I'd prefer this sum be denominated in steadily appreciating cash rather than constant deposits, and she will prefer the latter. What becomes the standard unit of account?


The last way to pay interest (at least as far as I know) is to run lotteries based on banknote serial numbers, an idea independently proposed by Hu McCulloch and Charles Goodhart back in 1986.

Central banks would periodically hold draws entitling the winning serial numbers to large cash prizes. For example, if there was $100 billion in banknotes in circulation, the central bank could set the interest rate on cash at 5% by offering prizes over the course of the year amounting to 5% of $100 billion, or $5 billion.

This technique of paying interest on cash solves the fungibility problem that plagues the earlier stamping technique. Every note has the same chance of winning the lottery, and non-fungible winners are immediately withdrawn. And unlike the crawling peg idea, banknotes and deposits remain equal to each other so burdensome exchange rate calculations don't need to me made.

However, it introduces the threat of bank runs. The day before the big lottery is set to occur, everyone will withdraw deposits for cash so that they can compete in the draw. To prevent a bank run, it may be necessary to randomize the date of the big lottery so that no one knows when to withdraw notes, an idea proposed by Tyler Cowen. Another way to preclude bank runs is to have a regular stream of small weekly lotteries rather than one or two big ones each year.

Another drawback to note lotteries is the cost that is imposed on society by having everyone constantly checking serial numbers. As Brian Romanchuk points out, employees who are working behind their employer's tills may be tempted to switch out winning notes with losers. Employers may protect themselves by setting up scanning hardware to read in serial numbers as banknotes enter the tills, maintaining their own internal database of cash inventories so that winners can quickly be isolated and returned. But all of that is costly. Would it be worth it?

Interestingly, there is some precedent for these sorts of lotteries. In Taiwan, receipts are eligible for a receipt lottery, a neat way to incentivize people to avoid under-the-table transactions (ht Gwern). Lotteries can also be useful in attracting depositors, as outlined in this Freakonomics podcast (ht Ryan). George Selgin and William Lastrapes have gone into the idea of lottery-linked money in some detail:
Though the suggestion may appear far fetched, in many countries lotteries are presently being used with considerable success to market bank deposits. According to Mauro Guillen and Adrian Tschoegl (2002), “lottery-linked” deposit accounts have been especially popular with poorer persons, including many who might otherwise remain “outside the banking system.” ... In two popular Argentine schemes, for instance, depositors receive one ticket or chance of winning for every $200 or $250 on deposit (ibid., p. 221). Lottery-linked banknotes, in contrast, would themselves serve as tickets, allowing persons to play for as little as the value of the lowest note denomination, and with no apparent cost to themselves save that of occasionally inspecting their note holdings.

Some readers may recognize these three techniques for paying interest on cash as the inverse of the three go-to ways of applying negative interest rates to cash being discussed a few years ago. For instance, one of the most well-known ways of imposing negative interest rates on owners of cash is to apply a Silvio Gesell style stamp scheme (see picture at top), whereby a currency owner must buy a stamp and affix it to the note in order to renew the validity of their currency each month. (I once discussed Alberta's experiment with Gesell's "shrinking money" here). Without the appropriate number of stamps, the note is illegitimate. In my first example above, Gesell's stamp tax has been re-engineered into a stamp subsidy. As for the magnetic strip modification, this is Marvin Goodfriend's 1999 update of Gesell, flipped around to award interest rather than docking it.

Miles Kimball has written extensively on escaping the zero lower bound to interest rates by setting a crawling peg on currency. But just as Kimball's crawling peg can impose a negative interest rate on banknotes, it can be used to pay interest, as I described above. Indeed, Miles (along with Ruchir Agarwal) frequently mention this possibility in his blog posts and papers (see this pdf).

Finally, remember Greg Mankiw's controversial 2009 article on imposing negative interest rates by serial number? He wrote:
Imagine that the Fed were to announce that, a year from today, it would pick a digit from zero to 9 out of a hat. All currency with a serial number ending in that digit would no longer be legal tender. Suddenly, the expected return to holding currency would become negative 10 percent.
Mankiw's idea is just the reverse of Goodhart and McCulloch's earlier lottery idea, the lottery replaced by with a demonetization.


So why pay interest on currency? I can think of two reasons. One is based on fairness, the other on efficiency.

The decision to avoid paying the market rate of interest on currency amounts to a tax on currency users. Who pays this tax? Cash is often the only means for the poor, new immigrants, and unbanked to participate in the economy. So the tax falls on those who can least afford it. This hardly seems fair. By conducting note lotteries or stamping notes, those consigned to the cash economy can get at least the same return on banknotes as the well-off banked receive on deposits.

Now hold up JP, some you will be saying at this point. What about criminals? Yep, the other group of people who suffer from the lack of interest on banknotes are criminals and tax evaders. Rewarding them with interest hardly seems appropriate. One would hope that if central banks did adopt a mechanism for rewarding currency with interest, it would be capable of screening out bad actors. For instance, criminals may be leery of collecting their interest or lottery prize if making a claim at a bank means potentially being unmasked. Another way to set up the screen would be to pay interest or prizes on small denominations like $1-$10 notes, and not on $20s and above. Since criminal organizations prefer high denomination notes due to their compactness, they wouldn't benefit from interest.

As for the efficiency argument, this is nothing but the famous Friedman rule that I described in my previous post. All taxes impose a deadweight loss on society. When a good or service is taxed, people produce and consume less of it than the would otherwise choose, tax revenues not quite compensating for this loss. From a policy maker's perspective, the goal is to reduce deadweight loss as much as possible by selecting the best taxes.

In the case of cash, the deadweight loss comes from people holding less of it than they would otherwise prefer, incurring so-called shoe leather costs as they walk to the bank and back to avoid holding too much of the stuff. If a 0% return on cash is an inefficient form of taxation relative to other alternatives types of taxes, then it would be better for the government to just pay interest on the stuff and recoup the lost revenues elsewhere, say through consumption taxes or income taxes.