It's time to trash the "store of value" function of money

When we first learn about money and banking in high school or university, we are all taught that money has three functions: medium of exchange, unit of account, and store of value. Maybe it’s time for educators to throw out this triumvirate. It’s not very accurate. 

We need a simple and teachable device to take the triumvirate’s place. I propose the money Venn diagram.

Before I explain the money Venn diagram, let’s revisit the textbook triumvirate.

When something is a medium of exchange, what is meant is that it is generally acceptable in trade. You can use it to buy stuff at the grocery store, or purchase stocks on the stock market, or get things online. 

The quality of being a medium of exchange is really more of a gradient than a matter of either/or. Banknotes, for instance, are good at brick and mortar shops, but useless online. Your debit card works great at shops, but forget trying to buy shares with it. But both are sufficiently widely accepted to qualify as a medium of exchange.

Because cryptocurrencies like Bitcoin and Litecoin aren’t widely accepted, they don’t make it across the line to qualify as a medium of exchange. Neither do Walmart or Target gift cards. Cigarettes don’t qualify either, but that wasn’t the case in 1950 when Milton Friedman used them to buy gas:

Milton Friedman describes the time he used cigarettes as money. (I get this from his book "Money Mischief".) pic.twitter.com/1NSjY9vZNX

— John Paul Koning (@jp_koning) November 6, 2020

The unit of account function of money refers to the fact that our economic conversations and calculations are couched in terms of a given monetary unit, whether that be the $, ¥, or £. In Canada and the US, prices are expressed in grocery aisles with dollars, our salaries use dollar units, and our debts are denominated in dollars. We don’t express prices in terms of government bonds, or Microsoft shares, or cigarettes or bitcoins. These things don’t function as a unit of account.

Thirdly, when money acts as a store of value we mean that it preserves value over time and space. Whereas the first two functions are quite useful, the store of value isn’t. Every asset functions as a store of value: houses, diamonds, banknotes, deposits, bitcoins, LSD tabs, lentils, cars, spices. And so it is meaningless to cast store of value as a unique function of money. Monetary economists such as Nick Rowe and George Selgin have proposed, and I concur, that we just chuck store of value from the definition of money.

But we are still left with two useful definitions for money, unit of account and medium of exchange. Which gets us to the money circle.

Note that the two circles in the diagram, medium of exchange and unit of account, don’t perfectly overlap. About 99% of the time the things we use as media of exchange are also the things we use as a unit of account. So the contents of our wallets or our bank accounts, dollar banknotes and dollar deposits are functionally equivalent to the $ units displayed in signs in grocery aisles.

But for the remaining 1% of the time, the unit of account and medium of exchange are separated. The idea of a separation is tough to get one’s head around. Luckily we’ve got a nice example. In Chile the prices of many things, particularly real estate, are expressed in terms of the Unidad de Fomento. But no Unidad de Fomento notes or coins circulate in Chile. It is a purely abstract unit of account.

Apartments for sale in Chile, priced in Unidad de Fomento

If a Chilean wants to buy an apartment that is priced at 840 Unidad de Fomento, she must use a separate medium of exchange, the Chilean peso, to make the payment. The peso is issued by Chile’s central bank, the Banco Central de Chile, in both paper and account form.

How many pesos must she pay? Every day the Banco Central de Chile publishes the exchange rate between the Unidad de Fomento and the peso. Right now one Unidad de Fomento is equal to 28,969 pesos. If an apartment were priced at 840 Unidad de Fomento, a Chilean would have to hand over 24 million Chilean pesos today.

Why has Chile separated its unit of account from its medium of exchange? I have discussed the issue at length. But the short answer is that it was a trick the government used to help cope with high inflation in the 1960s. Chilean inflation has been well under control for decades now. The practice of using the Unidad de Fomento as a unit-of-account has continued nonetheless.

You can see why it’s rare for these two functions to be separated. It’s awkward to do conversions every time one wants to pay for something. For the sake of ease, we tend to evolve towards systems where the medium of exchange and unit of account are united. But these exceptions are still important enough that we need a Venn diagram.

To sum up, money isn’t best thought of as a medium of exchange, unit of account, and store of value. Let’s just think of it as just a medium of exchange and a unit of account. For the most part these circles overlap, and the two functions are united. But this isn’t always the case. 

[My article was originally published at AIER's Sound Money Project in 2020 under the title A Simpler and More Accurate Way to Teach Money to Students]

The difference between two colourful bits of rectangular paper

David Andolfatto had a provocative and open-ended tweet a few days back:

The difference between money and debt. pic.twitter.com/CSQuLzUJPU

— David Andolfatto (@dandolfa) April 26, 2019

We see two coloured pieces of paper, both with an old dead President on it. They each have a face value of $500. Both are issued by a branch of the government, the $500 McKinley banknote (at right) by the Federal Reserve while the $500 Treasury bond (at left) by the Treasury. Both are bearer instrument: anyone can use them.

So why do we bestow one of them the special term "money" while the other is "credit"? I mean, they seem to be pretty much the same, right?

The word money is an awful word. It means so many different things to different people that any debate invoking the term is destined to go off-track within the first fifty characters. So I'm going to try and write this blog post without using the term money. Why are the two instruments that David has tweeted about fundamentally and categorically different from each other?

One of them is the medium of account, the other isn't

Being a veteran of the monetary economics blogosphere, David's tweet immediately made me think of the classical debates between Scott Sumner and Nick Rowe about the the medium-of-exchange vs medium-of-account functions of assets like banknotes and deposits and coins. (For those who don't remember, here are some posts.)

As Scott Sumner would probably say, one of the fundamental differences between the two bits of paper is that the McKinley $500 Federal Reserve note has been adopted as the U.S.'s medium-of- account. The $500 Treasury bond  hasn't.

Basically, if Jack is selling his car for $500, this price is represented by the $500 note (and other sub-denominations like the $50, $20 etc), not a $500 bond. Put differently, the bill is used as the medium for describing the accounting unit, the $. The bond does not have this special status. Now it could be that Jack is willing to accept bonds as payment, but since he doesn't use bonds to describe his sticker prices, he'll have to do some sort of calculation to convert the price into bond terms. When something is the medium of account, the entire language of prices is dictated by that instrument.

So why has society generally settled on using banknotes and not the bonds as our medium of account?

First, let's learn a bit more about the bond in question. The image that David has provided us with isn't actually a bond, it's a bond coupon. A coupon is a small ticket that the bond owner would periodically detach from the larger body of the bond in order to claim interest payments. The full bond would have looked more like this:

This format would have hobbled the bond's usefulness as a medium of exchange. The bond principal of $100 is represented by the largest sheet of paper. Attached to it are a bunch of coupons (worth $1.44 each) that haven't yet been stripped off. To compute the purchasing power of the bond, the $100 principal and all of the coupons would have to be added up. Complicating this summation is the time value of money. A coupon that I can clip off tomorrow is more valuable than the one I can clip off next year.

So if Jack is selling a car, and Jill offers him a $500 Treasury bond rather than a banknote, he'll have to spend a lot more time puzzling out the bond's value. A $500 McKinley note, which pays 0%, is much easier on the brain, and thus less likely to hit some sort of mental accounting barrier. (Larry White wrote a paper on this a while back).

Another hurdle is that there are many vintages of $500 Treasury bonds. A $500 bond that has been issued last year will be worth more than one that has been issued ten years ago and has had most of its coupons  stripped off. Put differently, Treasury bonds are not fungible. Banknotes, on the other hand, don't come in vintages. They are perfectly interchangeable with each other. So in places like stores and markets where trade must occur quickly, banknotes are far more convenient.

Further complicating matters is capital gains tax. Each time the $500 Treasury bond changes hands its owner must go back into their records to find the original price at which they received the bond, compute the profit, and then submit all this information to the tax authority. The $500 note doesn't face a capital gains tax. Better to use hassle-free banknotes, and not taxable bonds, to make one's day-to-day purchases.

Which finally gets us to why notes and not bonds are the medium-of-account. Since banknotes are such a convenient medium of exchange, everyone will have a few on hand. And this makes it convenient to set our prices in terms of notes, not Treasury bonds.

Why is it convenient? Say that Jack were to set the price of the car he is selling at $500, but tells his customers that the sticker price is in terms of Treasury bonds. So the $500 Treasury bond will settle the deal. But which Treasury bond does he mean? As I said earlier, at any point in time there are many vintages of $500 bonds outstanding. The 1945 one? The 1957 one? So confusing!

Jack's customers will all have a few notes in their wallet, having left their bonds locked away at home. But if Jack sets prices in terms of bonds, that means they'll have to make some sort of foreign exchange conversion back to notes in order to determine how many note to pay Jack. What a hassle!

If Jack sets the sticker price in terms of fungible notes he avoids the "vintages problem". And he saves the majority of his customers the annoyance of making a forex conversion from bond terms back into note terms. Since it's better to please customers than anger them, prices tend to be set in terms of the most popular payments instrument. Put differently, the medium-of-account tends to be married to the medium-of-exchange.

Alpha leaders vs beta followers

There is another fundamental difference between the two pieces of paper. Say that the Treasury were to adopt a few small changes to the instruments it issues. It no longer affixes coupons to Treasury bonds. And rather than putting off redemption for a few years, it promises to redeem them on demand with banknotes at any point in time. This new instrument would look exactly like the McKinley note. Without the nuisances of interest calculations, Treasury bond transactions should be just as effortless as the those with Federal Reserve notes.

But a fundamental difference between the two still exists. Since the Treasury promises to redeem the bond with banknotes, the Treasury is effectively pegging the value of the Treasury bond to the value of Federal Reserve notes. However, this isn't a reciprocal relationship. The Federal Reserve doesn't promise to redeem the $500 note with bonds (or with anything for that matter).

This means that the purchasing power of the bond is subservient to that of the banknote. Or as Nick Rowe tweets, "currency is alpha leader, bonds are beta follower."

Bonds promise to pay currency. Currency does not promise to pay bonds.
There's an asymmetric fixed (future) exchange rate between bonds and currency. Issuers of bonds peg to currency; issuers of currency do not peg to bonds.
So currency is alpha leader, bonds are beta follower.

— Nick Rowe (@MacRoweNick) April 26, 2019

This has much larger implications for the macroeconomy. In the long-run, the US's price level is set by the alpha leader, the Federal Reserve, not by the beta follower, the Treasury.

The Treasury could remove the peg. Now both instruments would be  0% floating liabilities of the issuer. Without a peg, their market values will slowly diverge depending on the policy of the issuer. For instance, a few years hence the $500 Treasury bond might be worth two $500 Federal Reserve notes. We could imagine that in certain parts of the U.S., custom would dictate a preference for one or the other as a medium of exchange. Or maybe legal tender laws nudge people into using one of them. And so certain regions would set price in terms of Treasury bonds while others will use Federal Reserve notes as the medium of account.

The Treasury's monetary policy would drive the price level in some parts of the U.S., whereas the Fed's monetary policy would drive it in the rest. This would be sort of like the 1860s. Most American states adopted Treasury-issued greenbacks as the medium of exchange during the Civil War, but California kept using gold coins issued by the US Mint. And thus prices in California continued to be described in terms of gold, and held steady, whereas prices in the East inflated as the Treasury printed new notes. (I wrote about this episode here.)

In conclusion...

So in sum, the two instruments in David's tweet are fundamentally and categorically different because one is the medium of account and the other isn't. Treasury bonds just aren't that easy to transact with, so people don't carry them around, and thus shopkeepers don't set sticker prices in terms of Treasury bonds. But even if the Treasury were to modify its bonds to be banknote look-alikes, they are still fundamentally different. Treasury paper is pegged to notes, but not vice versa.

This peg can be severed. But for convenience's sake, one of the two instruments will come to be used as the medium-of-account within certain geographical areas. And thus in its respective area, the issuer of that medium-of-account will dictate monetary policy.

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 power.

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:

Was Canada ground zero for fiat money? This paper claims that fiat money was independently invented by French Canadians in 1685, copied by Massachusetts in 1690, adopted in Europe, and diffused to the rest of the world: https://t.co/bSXyYLYzwp pic.twitter.com/kWVWbi43hH

— JP Koning (@jp_koning) March 14, 2018

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. 

Store of value

LSD tabs like these ones have an incredibly high value-to-weight ratio

When bitcoin first appeared, it was supposed to be used to buy stuff online. In his 2008 whitepaper, Satoshi Nakamoto even referred to his creation as an electronic cash system. But the stuff never caught on as a medium-of-exchange: it was too volatile, fees were too high, and scaling problems resulted in sluggish speeds. Despite losing its motivating purpose, bitcoin's price kept rising. The bitcoin cognoscenti began to cast around for a new raison d'etre. Invoking whatever they must have remembered from their old economics classes, they rechristened bitcoin as the world's best store of value.

Store of value is one of the three classic functions of money that we all learn about in Money and Banking 101: money serves a role as a medium of exchange, unit of account, and store of value. So presumably if bitcoin wasn't going to be a medium of exchange (and certainly not a unit of account thanks to its volatility), at least some claim to money-ishness could be retained by having it fill the store of value role.

In his 1867 Money and the Mechanism of Exchange, political economist William Stanley Jevons formally introduced the term store-of-value into monetary economics (although Nathan Tankus tells me that Marx may have originated the idea albeit with different terminology, and Daniel Plante tips Aristotle):

William Stanley Jevons (1867) was probably the first economist to introduce the 'store of value' function of money https://t.co/xqIs3Fym3a pic.twitter.com/dRcfLfMUya

— JP Koning (@jp_koning) October 8, 2017

Jevons's store of value function refers to the process of preserving value across both time and space. Now in one sense, every good that has ever existed has been a store of value, as Nick Rowe once pointed out. If a good isn't capable of storing value, we'd be incapable of handling and consuming it. Even an ice-cream cone needs to exist long enough for value to be transferred from tub to mouth.

What Jevons was implying in the above passage is that some goods are better than others at condensing value. Goods with the low bulk and weight, including the "current money of the land" (i.e. banknotes), are the best condensers. Below is a list of items ranked according to price per pound, which I get from Evilmadscientist (beware, these are 2008 prices). While all-purpose flour can store value, a $100 bill is better at the task, and while both are surpassed by championship thoroughbred semen, nothing does the job better than LSD.

To condense value over time and space, a store of value will need to be durable. Saffron has a fairly high value-to-weight ratio, but its quality depreciates much quicker than a dollar bill, thus compromising its ability to store value through time. Same with copper and silver, both of which will steadily corrode whereas gold does not. It also helps to have low storage costs. Oxen may have been a great way to store value across space, yet feeding and sheltering them over long periods of time would have been quite expensive. 

Jevons was writing before computers and the internet had emerged. Nowadays, billions of dollars in value are represented digitally. These digital tokens—stocks, bonds, deposits, credit, bitcoin, and whatnot—are weightless and volumeless. Which means they far exceed the ability of any physical item to condense value over time and space.

How does bitcoin rank relative to other digital stores of value? Let's say you needed to condense a certain amount of value and had a choice between either holding bitcoin or Netflix stock. (I choose Netflix because its market cap is close to the market value of all bitcoins ever mined, and because both their prices have done exceptionally well over the last six years). Bitcoin is great for conveying value across space, especially if it involves crossing national borders. All you have to do is remember your private key and you can access your funds no matter where you are. Netflix isn't quite so fluid. While you can certainly access your online brokerage account when you are in Vietnam on holiday, you can't actually sell Netflix stock in Vietnam (as you presumably could with bitcoin). Instead, you'd have to sell the stock and transfer the proceeds to a bank account in Vietnam via the correspondent banking system. That could take a few days and you might run into some hassles.

What about for storing value across time? Bitcoin has a few neat features, including censorship resistance. Since bitcoin isn't centrally managed, there is no way for an administrator to censor you, i.e. erase your bitcoins. With Netflix (or any other centrally-housed digital asset), however, if you are a considered to be a bad actor by those who control the system, presumably your shares can be frozen or confiscated. Counterbalancing this, bitcoins are notoriously susceptible to being stolen. But I've never heard of a thief getting away with someone's shares. There's a bit of give and take.

But in general, I'd argue that bitcoin and Netflix stock are both pretty bad for temporally storing value, although bitcoin is particularly bad. For an asset to do a good job condensing property over time it has to provide its owner with predictable access to a future basket of consumption goods. Assets with prices that have gone parabolic do not fulfill this requirement. After all, there is no reason that the price won't reverse and start to plunge, thus compromising that instrument's ability to store predictable amounts of consumption through time. Anything with a highly stable price across all time frames (minute-by-minute and year-to-year) provide the requisite predictability. Assets that gyrate do not.

The chart below shows the relative variability of the prices of bitcoin, Netflix, and gold since 2011.

Specifically, the chart measures each assets' median change in price over a given month. For instance, in November 2017 bitcoin had the tendency to close up or down by around 3.2% each day, Netflix by 0.8%, and gold by 0.3%. Averaging out all months since 2011, gold's variability comes in at 0.5%, Netflix at 1.5% and bitcoin at 2.2% (see dashed lines above), which means the yellow metal has done a much more predictable job of storing value over time than the other two assets, and Netflix is more up to the task than its digital counterpart. 

In late 2016 bitcoin's volatility seemed to have fallen permanently below Netflix levels and—for a month or two—approached that of gold. The digital stuff had become a mature asset! That wasn't to be, however, and bitcoin volatility has since reverted to levels significantly above its long term average.

I'd argue that bitcoin's high volatility is inherent to its nature. As such, it will always do a fairly bad job of storing value over time. The problem, as I outlined in my recent BullionStar article, is that bitcoin is a pure Keynesian beauty contest asset. People only buy bitcoins because they expect others to buy them at a higher price. The markets for gold and Netflix, on the other hand, are populated by a second set of participants who value those assets for reasons apart from whether others will buy them later. In the case of gold, industrial buyers step up whereas with Netflix it is value investors. The buying and selling of this second set of participants has a calming effect on prices.

The most predictable way to condense value through time is a U.S. dollar deposit. Anyone who has $100 in their account knows with a high degree of accuracy what they'll be able to buy next week. This stems from the fact that consumer good prices are measured in terms of the units issued by the central bank, and retailers keep these prices fairly rigid over the short term. For longer time periods, say one year out, the U.S. dollar will have naturally suffered from some inflation. But this decline in purchasing power is a known quantity. The Federal Reserve has an inflation target of 2%. So it's a safe bet that $100 will be worth $98, not $92, or $84, or $104. That's pretty good predictability. Interest earned on the deposits will make up for the lost purchasing power.

So is bitcoin a store of value? Sure, everything is to some degree... and bitcoin certainly does a good job of condensing value across distances. But relative to other assets, in particular U.S. dollar deposits, it does a poor job storing value across time. I don't think this is going to change, but I could be wrong.

---

P.S: In the interest of full disclosure, I still own some bitcoin and XRP, not much though. Own some gold too, but no Netflix.
P.P.S: Here is a rewrite of Satoshi's whitpaper, substituting in store of value system for electronic cash system:

Dictionary money

Nick Rowe points out that if a central bank wants to control the economy's price level, it needn't issue any actual money—it can just edit the dictionary every morning, announcing the meaning of the word "dollar" or "yen" or "pound" to the public.

To a modern ear trained on a steady diet of central bank verbiage about interest rates, QE, and open market operations, the idea of conducting monetary policy by simply editing the meaning of a word seems odd. But I've got news for you: starting from Caesar's time and extending into the 1700s, the sort of dictionary money that Nick describes has been the dominant form of money in the West.

How has this system worked? People have historically advertised prices for wares using a word, or unit of account, the LSD unit being the most prevalent. In the case of Britain this meant pound/shilling/pence while in France it was livre/sous/denier, both of which come from the Latin librae/solidi/denarii. The monarch was responsible for declaring what these words meant. More specifically, the king or queen would post a sign in some central area saying something to the effect that a pound, or £, was worth, say, ten testoons, a type of silver coin. This definition was subject to change. The next day, for instance, an edict might be issued saying that a £ was now only worth nine testoons. Or, put differently, the £ now contained less silver. Just like that, prices had to rise 10% to account for the alteration made to the dictionary meaning of the word "pound."

Dictionary systems came to an end when the symbol for money was finally fused directly with the instrument itself. Remember, coins never used to have denominations, or units of account, on their face. Rather, they usually only had the monarch's head inscribed on them, maybe the name of the mint, and a few words about how awesome the monarch was. This lack of numbering was convenient. Since coins had no association with the unit of account, the quantity of coins (and thus silver) in the unit of account (i.e. the definition of the word) could be seamlessly changed by royal proclamation.

In the 1700s monarchs began to adopt the practice of inscribing the actual unit of account directly on the coin's face, i.e. coins began to be etched with 5¢ or £0.5.* Once this happened it became awkward to change the definition of the unit of account by editing the dictionary. Having permanently stamped the meaning of the word "dollar" or "pound" on millions of widely-circulating bits of stamped silver, changing that meaning by simply posting a sign on a popular street corner no longer did the trick. Every coin would have to be recalled and re-minted too!

Having long since put the definition of the word "dollar" or "yen" onto the actual instruments they issue, modern monetary authorities now have to do something to the instruments themselves if they want to conduct monetary policy. Maybe they issue a few more units of money or buy them back in order to alter their purchasing power. Maybe they jiggle the interest rate that those tokens throw off. Or they might raise or lower a currency's peg. Some sort of tangible action (or threat thereof) must be taken to change the economy-wide price level. Word updates won't do.

About the only place in the world that has dictionary money is Chile which, buffeted by high inflation, adopted a parallel unit of account called the Unidad de Fomento (UF) in the 1960s. (For more on the UF, see my old post here). Today, Chileans can choose to set prices in UF or in the Chilean peso. The latter is a conventional money, the word "peso" being defined as the 1 peso banknote issued by the nation's central bank. Unlike the peso, the UF lacks an underlying UF banknote. Rather, the Chilean government defines the word "Unidad de Fomento" to mean the number of Chilean pesos required to buy a fixed Chilean consumption basket. This definition changes every day and is posted here.

I think this is a pretty neat idea. As long as Chileans denominate their salary and other contracts using UFs rather than pesos, they are guaranteed to earn a steady stream of consumption, even if the Chilean peso hyperinflates.

These days inflation isn't really such a big deal, at least not in developed nations—central bankers seem to have mastered how to keep the purchasing power of the medium of exchange from getting out of hand. So adopting something like the UF might seem redundant. A dictionary money system is also unattractive because it imposes a calculational burden on citizens. People must be constantly doing conversions between an item's sticker price and whatever happens to be the medium of exchange necessary to complete the transaction. So if a book were to be priced at $5, you'd have to consult a government website to determine how many bitcoins, or dollar bills, or silver coins would be necessary to constitute a five dollar payment. The advantage of our current system is that because the word and the medium are unified, we don't have to do these conversions. A five dollar bill always suffices to cover a $5 sticker price. Simple.

On the other hand, dictionary money may have a role to play in our relatively recent deflationary age. Beginning with Japan back in the late 1990s, central bankers all over the world have been incapable of preventing deflation, or falling prices. Are their tools inadequate? Do they refuse to use these tools to their full extent? Do they not understand how to use them? With dictionary money, a central banker can't blame his or her tools for a miss, since all it takes to alter the price level is an update to the definition. A child could do it.

For instance, a nation like Japan could create dictionary money by removing the word "yen" on bills. It would do so by recalling all outstanding banknotes and replacing them with, say, Japanese pesos. Prices, however, would continue to be set in terms of the yen unit of account. Each morning the Bank of Japan would announce to the world how many Japanese pesos were in a yen. Say it starts with the yen being defined as ten pesos. To create some inflation, it would simply proclaim that the yen now contained just five pesos. Everyone with pesos in their pocket would suddenly be able to buy twice as much yen-denominated products as before. They would race out and spend. Shopkeepers who had previously been selling widgets for 1 yen, and getting ten Japanese pesos as payment, would quickly jack up prices to 2 yen in order to ensure that they still earn ten pesos per widget.

Voila, instant inflation.


* See Ernst Weber's "Pre-industrial Bimetallism: The Index Coin Hypothesis " [link]

Kocherlakota on cash

Narayana Kocherlakota, formerly the head of the Federal Reserve Bank of Minneapolis and now a prolific economics blogger, penned a recent article on the abolition of cash. Kocherlakota makes the point that if you don't like government meddling in the proper functioning of free markets, then you shouldn't be a big fan of central bank-issued banknotes. For markets to clear, it may be occasionally necessary for nominal interest rates to fall well below zero. Cash sets a lower limit to interest rates, thus preventing this rebalancing from happening.

I pretty much agree with Kocherlakota's framing of the point. In fact, it's an angle I've taken before, both here and in A Libertarian Case for Abolishing Cash. Yes, my libertarian and other free-marketer readers, you didn't misread that. There is a decent case for removing banknotes that is entirely consistent with libertarian principles. If you think usury laws are distortionary because they impose a ceiling on interest rates—and there are some famous libertarians who have railed against usury—then an appeal to symmetry says that you should be equally furious about the artificial, and damaging, interest rate floor set by cash.

Scott Sumner steps up to the plate and defends cash here. He brings up some good points, but I'm going to focus on his last one. Scott says that a cashless economy would create a "giant panopticon" where the state knows everything about you. I quite like Nick Rowe's response in which he welcomes Scott to the Margaret Atwood Club for the Preservation of Currency. In Atwood's dystopian Handmaid's Tale, a theocratic government named the Republic of Gilead has taken away many of the rights that women currently enjoy. One of the tools the Republic uses to control women is a ban on cash, all transactions now being routed digitally through something called the Compubank:

I agree that we don't want to abolish cash if it is only going to lead to Atwood's Compubank. But Scott misses the fact that even though Kocherlakota wants the government to exit the cash business, he simultaneously wants fintech companies to take up the mantle of anonymity services provider. Like Sumner, Kocherlakota doesn't seem to want a Compubank.

For instance, in a recent presentation entitled The Zero Lower Bound and Anonymity: A Monetary Mystery Tour, Kocherlakota highlights the potential for cryptocoins Zcash and Monero to substitute for central bank cash. Unlike bitcoin, these cryptocoins provide full anonymity rather than just pseudonymity. If you want to learn more about Zcash, I just listened to a great podcast with Zcash's Zooko Wilcox-O'Hearn here. As for Monero, Bloomberg recently covered its spectacular rise in price.

As Monero illustrates, cryptocoins are incredibly volatile. Is anonymity too important of a good to be outsourced to assets that behave like penny stocks? I'm not sure. And as Nick Rowe points out, the concurrent circulation of deposits (pegged to central bank money) and anonymity-providing cryptocoins would create havoc with the traditional way of accounting for prices. Retailers would probably still set prices in terms of central bank money but anyone wanting to purchase something anonymously would have to engage in an inconvenient ritual of exchange rate conversion prior to consummating the deal. Perhaps these are simply the true costs of enjoying anonymity?

Kocherlakota doesn't mention it explicitly, but should cash be abolished in order to remove the lower bound to interest rates, a potential replacement would be a new central bank-issued emoney, either Fedcoin or what Dave Birch has dubbed FedPesa. A good example of a Fedcoin-in-the-works comes from the People's Bank of China, which vice governor Fan Yifei expects to "gradually replace paper money." As for Birch's FedPesa, a real life example of this is provided by Ecuador's Dinero electrónico, a mobile money scheme maintained by the Central Bank of Ecuador (CBE) for use by the public.

Should a government decide to abolish cash and implement a central bank emoney scheme in its place, it would be possible to set negative interest rates on these tokens while at the same time promising to provide both stability and anonymity. One wonders how credible the latter promise would be. The CBE requires that citizens provide national identity card before opening accounts. And consider that the PBoC's potential cyptocoin will be designed to provide "controlled anonymity," whatever that means. Unless significant safeguards are set, it's hard not to worry that a potential Atwood-style Compubank is waiting in the wings.

An alternative way to coordinate a smooth government exit from the cash business is Bill Woolsey's idea of allowing private banks to step into the role of providing banknotes. In this scenario, the likes of HSBC, Bank of America, Wells Fargo, Deutsche Bank, and Royal Bank of Canada would become sole providers of circulating banknotes. Wouldn't this simply re-establish the zero lower bound? Not necessarily. As I wrote back in 2013, the moment a central bank sets deeply negative interest rates, private banks will face huge incentives to either 1. get out of the business of cash or 2. stay in the game while modifying arrangements, the effect being that the zero lower bound is quickly ripped apart.

The provision of anonymity services via the issuance of private banknotes has some advantages over cryptocoins like Zcash. Since they'd be pegged to central bank money, private banknotes would provide 'fixed-price' anonymity. Nor would the public have to constantly do exchange rate conversions between one currency type or the other. On the other hand, Zcash payments can be made instantaneously over long distances; you just can't do that with banknotes. And of course, there's also the stablecoin dream, i.e. the possibility that private cryptocoins like Zcash might themselves be stabilized by pegging them to central bank cash, as Will Luther describes here (for a more skeptical take, read R3's Kathleen B here)

Because of what he calls "over-issue" problems, Kocherlakota is more confident in the prospects for cryptocoins than private banknotes. I'm not so worried. The voluminous free-banking literature developed by people like George Selgin, Larry White, and Kevin Dowd teaches us that as long as silly regulations are avoided, the promise to redeem notes at par in a competitive environment will ensure that the quantity of private banknotes supplied never exceeds the quantity demanded. Don't look to the so-called U.S. Wildcat banking era for proof. During that era, note-issuing banks were too encumbered by strict laws against branch banking and cumbersome backing rules to effectively supply notes, as Selgin points out here. Rather, the Scottish and Canadian banking systems of the 1800s provide evidence that banks can responsibly issue paper money.

Wouldn't the private provision of banknotes require the passing of new laws? Funny enough, U.S. commercial banks can already issue their own banknotes. In a fascinating 2001 article, Kurt Schuler points out that federally-chartered banks have been free to issue notes since 1994 when restrictions on note issuance by national banks was repealed as obsolete by the Community Development Banking and Financial Institutions Act. So the floodgates are open, in the U.S. at least, although as of yet no bank has taken the lead.

If governments are going to remove the zero lower bound by getting out of the business of providing anonymous payments, I say let a thousand flowers bloom. If the void is to be filled, don't put up any impediments to the creation of anonymity-providing fintech options like Zcash, but likewise don't prevent old fashioned banks from getting into the now-vacated banknote game either. Let the market decide which anonymity product they prefer... and celebrate the fact that the government's artificial floor to interest rates has been dismantled.

---

P.S. It would be remiss of me to omit pointing out that there are sound ways to dismantle the zero lower bound without removing cash, Miles Kimball's plan being one of them.

Why hasn't Canadian Tire Money displaced the Canadian dollar?

Canadians will all know what Canadian Tire Money is, but American and overseas readers might not. Canadian Tire, one of Canada's largest retailers, defies easy categorization, selling everything from tents to lawn furniture to hockey sticks to car tires. Since 1958, it has been issuing something called Canadian Tire Money (see picture above). These paper notes are printed in denominations of up to $2 and are redeemable at face value in kind at any Canadian Tire store.

Because there's a store in almost every sizable Canadian town, and the average Canadian make a couple visits each year, Canadian Tire money has become ubiquitous—everyone has some stashed in their cupboard somewhere. Many Canadians are quite fond of the stuff—there's even a collectors club devoted to it. I confess I'm not a big fan: Canadian Tire money is form of monetary pollution, say like bitcoin dust or the one-cent coin. I just throw it away.

It's the monetary oddities that teach us the most about monetary phenomena, which is why I find Canadian Tire money interesting. Here's an observation: despite the fact that it is ubiquitous, looks like money, trades at par, and is backed by a reputable issuer, Canadian Tire money doesn't circulate much. Stephen Williamson, a Canadian econ blogger, had an entertaining blog post a few years back recounting unsuccessful efforts to offload the coupons on Canadians. Sure, from time to time we might encounter the odd bar or charity that accepts it, or maybe a corner-store in Wawa. But apart from Canadian Tire stores, acceptability of Canadian Tire money is the exception, not the rule.

Why hasn't Canadian Tire money become a generally-accepted medium of exchange? One explanation is that Canadian law prevents it. Were the government to remove the strict rules that limit the ability of the private sector to issue paper money, bits of Canadian Tire paper would soon be circulating all across the nation, maybe even displacing the Bank of Canada's paper money.

A second hypothesis is that even if the law were to be loosened, Canadian Tire would remain an unpopular exchange medium. Some deficiency with Canadian Tire money, and not strict laws, drives their lack of liquidity. Nick Rowe, another Canadian econ blogger (notice a theme here?), once speculated that this had to do with network effects. Canadians have long since adopted the convention of using regular Bank of Canada-issued notes, and overturning that convention by accepting Canadian Tire money would be too costly. David Andolfatto (not another Canadian econ blogger!), would probably point to limited commitment as the deficiency. IOUs issued by Canadian Tire simply can't be trusted as much as government money, and so they inevitably fail as a medium of exchange.

In support of the first view, which is known in the economics literature as the legal restrictions hypothesis, Neil Wallace and Martin Eichenbaum (yep, another Canadian) recount an interesting anecdote. Back in 1983, competitor Ro-Na (since renamed RONA), a major hardware chain, started to accept Canadian Tire coupons at face value. I've found an old advertisement of the offer below:

RONA advertisement in La Presse, 1983 (source)

Eichenbaum and Wallace say that this is evidence that Canadian Tire money isn't just a mere coupon but readily serves as a competitive medium of exchange among Canadians. After all, if a major store like RONA accepted the coupons, then their acceptance wasn't just particular—it was general.

The story doesn't end there. Here is an interesting 1983 article from the Montreal Gazette:

The article mentions how in retaliation Canadian Tire sought an injunction against RONA to prevent it from accepting Canadian Tire money. We know this tactic must have been somewhat successful since RONA does not currently accept said coupons. Eichenbaum & Wallace slot this into their legal restrictions theory, noting that the injunction was probably motivated by Canadian Tire's desire to comply with legal prohibitions on the private issuance of currency, a damaging law suit helping to inhibit general use of their coupons. Remove these legal restrictions, however, and Canadian Tire would probably not have sued RONA, and usage of Canadian Tire coupons as a medium of exchange would have expanded. Presumably if Tim Horton's took up the baton from RONA and accepted Canadian Tire money, and then Couche-Tard joined in, you'd end up with a new national currency.

So we have two competing theories to explain Canadian Tire money's lack of acceptability. Which one is right? Let's introduce one more story arc. Zoom forward to 2009 when Canadian Tire lawyers sent a notice to a NAPA car parts dealer asking him to stop accepting Canadian Tire money. The reason cited by Canadian Tire: trademark infringement. As the article points out, Canadian Tire Money constitutes intellectual property, and if companies do not sufficiently police their trademarks against general usage, they may lose control of them. For instance, over the years Johnson & Johnson has had to vigorously defend its exclusive rights to the name "Band-Aid." If it hadn't, it might have lost claim to the name in the same way that Otis Elevator lost its trademark on the word "escalator" because the word fell into general use. That the 1983 RONA challenge probably had less to do with currency laws than trademark infringement damages Eichenbaum &Wallace's argument.

The last interesting Canadian factoid is the observation that a number of community currencies circulate in Canada. Salt Spring dollars, a currency issued by the Salt Spring Island Monetary Foundation, located off the coast of British Columbia, is one of these. Other examples include Calgary Dollars and Toronto Dollars. According to Johanna McBurnie, Salt Spring dollars are legal because they are classified as gift certificates. If so, I don't see why the use of Canadian Tire money as a medium of exchange wouldn't fall under the same rubric. This puts the final nail in Wallace & Eichenbaum's argument that restrictions on circulation of competing paper money have prevented broad usage of Canadian Tire paper. Rather, if laws are to blame for the minimal role of Canadian Tire Money's as currency, then it is the company's desire to protect its trademark that is at fault.

That local IOUs like Salt Spring dollars can legally circulate but lack wide acceptance (even in the locality in which they are issued) means we need something like the Nick Rowe's network effects or David Andolfatto's limited commitment to  explain why incumbent paper money tends to exclude competing paper money from circulation. Which isn't to say that Canadian Tire money would never circulate. As Larry White and George Selgin have pointed out, private paper money has circulated along with government paper money in places like Canada. But the bar for Canadian Tire money is probably a high one.

A 21st century gold standard

Imagine waking up in the morning and checking the hockey scores, news, the weather, and how much the central bank has adjusted the gold content of the dollar overnight. This is what a 21st century gold standard would look like.

Central banks that have operated old fashioned gold standards don't modify the gold price. Rather, they maintain a gold window through which they redeem a constant amount of central bank notes and deposits with gold, say $1200 per ounce of gold, or equivalently $1 with 0.36 grains. And that price stays fixed forever.

Because gold is a volatile commodity, linking a nation's unit of account to it can be hazardous. When a mine unexpectedly shuts down in some remote part of the world, the necessary price adjustments to accommodate the sudden shortage must be born by all those economies that use a gold-based unit of account in the form of deflation. Alternatively, if a new technology for mining gold is discovered, the reduction in the real price of gold is felt by gold-based economies via inflation.

Here's a modern fix that still includes gold. Rather than redeeming dollar bills and deposits with a permanently fixed quantity of gold, a central bank redeems dollars with whatever amount of gold approximates a fixed basket of consumer goods. This means that your dollar might be exchangeable for 0.34 grains one day at the gold window, or 0.41 the next. Regardless, it will always purchase the same consumer basket.

Under a variable gold dollar scheme the shuttering of a large gold mine won't have any effect on the general price level. As the price of gold begins to skyrocket, consumer prices--the reciprocal of a gold-linked dollar--will start to plummet. The central bank offsets this shock by simply redefining the dollar to contain less gold grains than before. With each grain in the dollar more valuable but the dollar containing fewer grains of the yellow metal, the dollar's intrinsic value remains constant. This shelters the general price level from deflation.

This was Irving Fisher's 1911 compensated dollar plan  (see chapter 13 of the Purchasing Power of Money), the idea being to 'compensate' for changes in gold's purchasing power by modifying the gold content of the dollar. A 1% increase in consumer prices was to be counterbalanced by a ~1% increase in the number of gold grains the dollar, and vice versa. Fisher referred to this fluctuating definition as the 'virtual dollar':

From A Compensated Dollar, 1913

Fisher acknowledged that 'embarrassing' speculation was one of the faults of the system. Say the government's consumer price report is to be published tomorrow and everyone knows ahead of time that the number will show that prices are rising too slow. And therefore, the public expects that the central bank will have to increase its gold buying price tomorrow, or, put differently, devalue the virtual dollar so it is worth fewer ounces of gold. As such, everyone will rush to exchange dollars for gold at the gold window ahead of the announcement and sell back the gold tomorrow at the higher price. The central bank becomes a patsy.

Fisher's suggested fix  was to introduce transaction costs, namely by setting a wide difference between the price at which the central bank bought and sold gold. This would make it too expensive buy gold one day and sell it the next. This wasn't a perfect fix because if the price of gold had to be adjusted by a large margin the next day in order to keep prices even, say because a financial crisis had hit, then even with transaction costs it would still be profitable to game the system.

A more modern fix would be to adjust the gold content of the virtual dollar in real-time in order to remove the window of opportunity for profitable speculation. Given that consumer prices are not reported in real-time, how can the central bank arrive at the proper real-time gold price? David Glasner once suggested targeting the expectation. Rather than aiming at an inflation target, the central bank targets a real-time market-based indicator of inflation expectations, say the TIPS spread. So if inflation expectations rise above a target of 2% for a few moments, a central bank algorithm rapidly reduces its gold buying price until expectations fall back to target. Conversely, if expectations suddenly dip below target, over the next few seconds the algorithm will quickly ratchet down the content of gold in the dollar to whatever quantity is sufficient to restore the target (i.e. it increases the price of gold).

Gold purists will complain that this is a gold standard in name only. And they wouldn't be entirely wrong. Instead of defining the dollar in terms of gold, a compensated dollar scheme could just as well define it as a varying quantity of S&P 500 ETF units, euros, 10-year Treasury bonds, or any other asset. No matter what instrument is being used, the principles of the system would be the same.

A compensated dollar scheme isn't just a historical curiosity; it may have some relevance in our current low-interest rate environment. Lars Christensen and Nick Rowe have pointed out that one advantage of Fisher's plan is that it isn't plagued by the zero lower bound problem. Our current system depends on an interest rate as its main tool for controlling prices. But once the interest rate that a central bank pays on deposits has fallen below 0%, the public begins to convert all negative-yielding deposits into 0% yielding cash. At this point, any further attempt to fight a deflation with rate cuts is not possible. The central banker's ability to regulate the purchasing power of money has broken down.*

Under a Fisher scheme the tool that is used to control purchasing power is the price of gold, or the gold content of the virtual dollar, not an interest rate. And since the price of gold can rise or fall forever (or alternatively, a dollars gold content can always grow or fall), the scheme never loses its potency.

Ok, that is not entirely correct. In the same way that our modern system can be crippled under a certain set of circumstances (negative rates and a run into cash), a Fisherian compensated dollar plan had its own Achilles heel. If gold coins circulate along with paper money and deposits, then every time the central bank reduces the gold content of the virtual dollar in order to offset deflation it will have to simultaneously call in and remint every coin in circulation in order to keep the gold content of the coinage in line with notes and deposits. This series of recoinages would be a hugely inconvenient and expensive.

If the central bank puts off the necessary recoinage, a compensated dollar scheme can get downright dangerous. Say that consumer prices are falling too fast (i.e. the dollar is getting too valuable) such that the central banker has to compensate by reducing the gold content of the virtual dollar from 0.36 grains to 0.18 grains (I only choose such a large drop because it is convenient to do the math). Put differently, it needs to double the gold price to $2800/oz from $1400. Since the central bank chooses to avoid a recoinage, circulating gold coins still contain 0.36 grains.

The public will start to engage in an arbitrage trade at the expense of the central bank that goes like this: melt down a coin with 0.36 grains and bring the gold bullion to the central bank to have it minted into two coins, each with 0.36 grains (remember, the central bank promises to turn 0.18 grains into a dollar, whether that be a dollar bill, a dollar deposit, or a dollar coin, and vice versa). Next, melt down those two coins and take the resulting 0.72 grains to the mint to be turned into four coins. An individual now owns 1.44 grains, each coin with 0.36 grains. Wash and repeat. To combat this gaming of the system the government will declare the melting-down of  coin illegal, but preventing people from running garage-based smelters would be pretty much impossible. The inevitable conclusion is that the public increases their stash of gold exponentially until the central bank goes bankrupt.

This means that a central bank on a compensated dollar that issues gold coins along with notes/deposits will never be able to fight off a deflation. After all, if it follows its rule and reduces the gold content of the virtual dollar below the coin lower bound, or the number of grains of gold in coin, the central bank implodes. This is the same sort of deflationary impotence that a modern rate-setting central bank faces in the context of the zero lower bound to interest rates.

In our modern system, one way to get rid of the zero lower bound is to ban cash, or at least stop printing it. Likewise, in Fisher's system, getting rid of gold coins (or at least closing the mint and letting existing coin stay in circulation) would remove the coin lower bound and restore the potency of a central bank. Fisher himself was amenable to the idea of removing coins altogether. In today's world, the drawbacks of a compensated dollar plan are less salient as gold coins have by-and-large given way to notes and small base metal tokens.

In addition to evading the lower bound problem, a compensated dollar plan would also be better than a string of perpetually useless quantitative easing programs. The problem with quantitative easing is that commitments to purchase, while substantial in size, are not made at any particular price, and therefore private investors can easily trade against the purchases and nullify their effect. The result is that the market price of assets purchased will be pretty much the same whether QE is implemented or not. Engaging in QE is sort of like trying to change the direction of the wind by waving a flag, or, as Miles Kimball once said, moving the economy with a giant fan. A compensated dollar plan directly modifies the price of gold, or, alternatively, the gold content of the dollar, and therefore has an immediate and unambiguous effect on purchasing power. If central bankers adopted Fisher's plan, no one would ever accuse them of powerlessness again.

---

*Technically, interest rates need never lose their potency if Miles Kimball's crawling peg plan is adopted. See here.

Central banks' shiny new tool: cash escape inhibitors

Thomas Jordan, Chairman of the Governing Board of the Swiss National Bank

Negative interests rates are the shiny new thing that everyone wants to talk about. I hate to ruin a good plot line, but they're actually kind of boring; just conventional monetary policy except in negative rate space. Same old tool, different sign.

What about the tiering mechanisms that have been introduced by the Bank of Japan, Swiss National Bank, and Danmarks Nationalbank? Aren't they new? The SNB, for instance, provides an exemption threshold whereby any amount of deposits that a bank holds above a certain amount is charged -0.75% but everything within the exemption incurs no penalty. As for the Bank of Japan, it has three tiers: reserves up to a certain level (the 'basic balance') are allowed to earn 0.1%, the next tier earns 0%, and all remaining reserves above that are docked -0.1%.

But as Nick Rowe writes, negative rate tiers—which can be thought of as maximum allowed reserves—are simply the mirror image of minimum required reserves at positive rates. So tiering isn't an innovation, it's just the same old tool we learnt in Macro 101, except in reverse.

No, the novel tool that has been created is what I'm going to call a cash escape inhibitor.

Consider this. When central bank deposit rates are positive, banks will try to minimize storage of 0%-yielding banknotes by converting them into deposits at the central bank. When rates fall into negative territory, banks do the opposite; they try to maximize storage of 0% banknote storage. Nothing novel here, just mirror images.

But an asymmetry emerges. Central bankers don't care if banks minimize the storage of banknotes when rates are positive, but they do care about the maximization of paper storage at negative rates. After all, if banks escape from negative yielding central bank deposits into 0% yielding cash, this spells the end of monetary policy. Because once every bank holds only cash, the central bank has effectively lost its interest rate tool.

If you really want to find something innovative in the shift from positive to negative rate territory, it's the mechanism that central bankers have instituted to inhibit the combined threat of mass paper storage and monetary policy impotence. Designed by the Swiss and recently adopted by the Bank of Japan, these cash escape inhibitors have no counterpart in positive rate land.

The mechanics of cash escape inhibitors

Cash escape inhibitors delay the onset of mass paper storage by penalizing any bank that tries to replace their holdings of negative yielding central bank deposits with 0%-yielding cash. The best way to get a feel for how they work is through an example. Say a central bank has issued a total of $1000 in deposits, all of it held by banks. The central bank currently charges banks 0% on deposits. Let's assume that if banks choose to hold cash in their vaults they will face handling & storage costs of 0.9% a year.

Our central bank, which uses tiering, now reduces deposit rates from 0% to -1%. The first tier of deposits, say $700, is protected from negative rates, but the second tier of $300 is docked 1%, or $3 a year. Banks can improve their position by converting the entire second tier, the penalized portion of deposits, into cash. Each $100 worth of deposits that is swapped into cash results in cost savings of 10 cents since the $0.90 that banks will incur on storage & handling is an improvement over the $1 in negative interest they would otherwise have to pay. Banks will very rapidly withdraw all their tier-2 deposits, monetary impotence being the result.

To avoid this scenario, central banks can install a Swiss-style cash escape inhibitor. The way this mechanism works is that each additional deposit that banks convert into vault cash reduces the size of the first tier, or the shield, rather than the second tier, the exposed portion. So when rates are reduced to -1%, should banks try to evade this charge by converting $100 worth of deposits into vault cash they will only succeed in reducing the protected tier from $700 to $600, the second tier still containing the same $300 in penalized deposits. This evasion effort will only have made banks worse off. Not only will they still be paying $3 a year in negative interest but they will also be incurring an extra $0.90 in storage & handling ($100 more in vault cash x 0.9% storage costs).

Continuing on, if the banks convert $200 worth of deposits into vault cash in order to avoid -1% interest rates, they end up worsening their position even more, accumulating $1.80 in storage & handling costs on top of $3.00 in interest. We can calculate the net loss that the inhibitor imposes on banks for each quantity of deposits converted into vault cash and plot it:

The yearly cost of holding various quantities of cash at a -1% central bank deposit rate

Notice that the graph is kinked. When a bank has replaced $700 in deposits with cash, additional cash withdrawals actually reduce its costs. This is because once the first tier, the $700 shield, is used up, the next deposit conversion reduces the second tier, the exposed portion, and thus absolves the bank of paying interest costs. And since interest costs are larger than storage costs, overall costs decline.

If banks go all-out and cash in the full $1000 in deposits, this allows them to completely avoid the negative rate penalty. However, as the chart above shows, storage & handling costs come out to $9 per year ($1000 x 0.9%), much more than the $3 banks would bear if they simply maintained their $300 position in -1% yielding deposits.

So at -1% deposit rates and with a fully armed inhibitor installed, banks will choose the left most point on the chart—100% exposure to deposits. Mass cash conversion and monetary policy sterility has been avoided.

How deep can rates go?

How powerful are these inhibitors? Specifically, how deep into negative rate territory can a central bank go before they start to be ineffective?

Let's say our central banker reduces deposit rates to -2%. Banks must now pay $6 a year in interest ($300 x 2%). If banks convert all $1000 in deposits into cash, they will have to bear $9 in storage and handling costs, a more expensive option than remaining in deposits. So even at -2% rates, the cash inhibitor mechanism performs its task admirably.

If the central bank ratchets rates down to -3%, banks will now be paying $9 a year in interest ($300 x 3%). If they convert all $1000 in deposits into cash, they'll have to pay $9 in storage & handling. So at -3%, bankers will be indifferent between staying invested in deposits or converting into cash. If rates go down just a bit more, say to -3.1%, interest costs are now $9.30. A tipping point is reached and cash will be the cheaper option. Mass cash storage ensues, the cash escape inhibitor having lost its effectiveness.

The chart below shows the costs faced by banks at various levels of cash holdings when rates fall to -3%. The extreme left and right options on the plot, $0 in cash or $1000, bear the same costs.

The yearly cost of holding various quantities of cash at a -3% central bank deposit rate

So without an inhibitor, the tipping point for mass cash storage and monetary policy impotence lies at -0.9%, the cost of storing & handling cash. With an inhibitor installed the tipping point is reduced to -3.1%. The lesson being that cash escape inhibitors allow for extremely negative interest rates, but they do run into a limit.

The exact location of the tipping point is sensitive to various assumptions. In deriving a -3.1% escape point, I've used what I think is a reasonable 0.9% a year in storage and handling costs. But let's assume these costs are lower, say just 0.75%. This shifts the cash tipping point to around -2.5%. If costs are only 0.5%, the tipping point rises to around -1.7%.

This is where the size of note denominations is important. The Swiss issue the 1000 franc note, one of the largest denomination notes in the world, which means that Swiss cash storage costs are likely lower than in other countries. As such, the Swiss tipping point is closer to zero then in countries like the Japan or the U.S.. One way to push the tipping point further into negative terriotry would be a policy of embargoing the largest note. The central bank, say the SNB, stops printing new copies of its largest value note, the 1000 fr. Banks would no longer be able to flee into anything other than small value notes, raising their storage and handling costs and impinging on the profitability of mass cash storage.

Good old fashioned financial innovation will counterbalance the authorities attempts to drag the tipping point deeper. Cecchetti & Shoenholtz, for instance, have hypothesized that in negative rate land, a new type of intermediary could emerge that provides 'cash reserve accounts.' These specialists in cash storage would compete to reduce the costs of keeping cash, pushing the tipping point back up to zero.

The tipping point is also sensitive to the size of the first tier, or the shield. I've assumed that the central bank protects 70% of deposits from the negative deposit rate. The larger the exempted tier the bigger the subsidy central banks are providing banks. It is less advantageous for a bank to move into cash when the subsidy forgone is a large one. So a central bank can cut deeper into negative territory the larger the subsidy. For instance, using my initial assumptions, if the central bank protects 80% of deposits, then it can cut its deposit rate to -4.6% before mass paper storage ensues.

Removing the tipping point?

There are ways to modify these Swiss-designed cash escape inhibitors to remove the tipping point altogether. The way the SNB and BoJ have currently set things up, banks that try to escape negative rates only face onerous penalties on cash conversions as long as the first tier, the shield, has not been entirely drawn down. Any conversion after the first tier has been used up is profitable for a bank. That's why the charts above are kinked at $700.

If a central bank were to penalize cumulative cash withdrawals (rather than cash withdrawals up to a fixed ceiling) then it will have succeeded in snipping away the tipping point. This is an idea that Miles Kimball has written about here. One way to implement this would be to require that the tier 1 exemption, the shield, go negative as deposits continue to be converted into cash, imposing an obligation on banks to pay interest. The SNB doesn't currently allow this; it sets a lower limit to its exemption threshold of 10 million francs. But if it were to remove this lower limit, then it would have also removed the tipping point.

What about retail deposits?

You may have noticed that I've left retail depositors out of this story. That's because the current generation of cash escape inhibitors is designed to prevent banks from storing cash, not the public.

As central bank deposit rates fall ever deeper into negative territory, any failure to pass these rates on to retail depositors means that bank margins will steadily contract. If banks do start to pass them on, at some point the penalties may get so onerous that a run develops as retail depositors start to cash out of deposits. The entire banking industry could cease to exist.

To get around this, the FT's Martin Sandbu suggests that banks could simply install cash escape inhibitors of their own. Miles Kimball weighs in, noting that banks may start applying a fee on withdrawals, although his preferred solution is a re-deposit fee managed by the central bank. Either option would allow banks to preserve their margins by passing negative rates on to their customers.

Even if banks don't adopt cash escape inhibitors of their own, I'm not too worried about retail deposit flight in the face of negative central bank deposit rates of -3% or so. The deeper into negative rate territory a central bank progresses, the larger the subsidy it provides to banks via its first tier, the shield.  This shielding can in turn be transferred by a bank to its retail customers in the form of artificially slow-to-decline deposit rates. So even as a central bank reduces its deposit rate to -3% or so, banks might never need to reduce retail deposit rates below -0.5%. Given that cash handling & storage costs for retail depositors are probably about the same as institutional depositors, banks that set a -0.5% retail deposit rate probably needn't fear mass cash conversions.

So there you have it. Central banks with cash escape inhibitors can get pretty far into negative rate land, maybe 3% or so. And with a few modifications they might be able to go even lower.