Saturday, February 13, 2016

Can a central bank eliminate its highest value banknote?

Singapore's $10,000 bill, worth around US$7500, shares title to world's largest value banknote with Brunei's $10,000

Peter Sands has adeptly made the case for eliminating high denomination banknotes. The rough idea is that if all central banks were to eliminate their highest value banknotes, then criminals would have to fall back on smaller denominations or more volatile media of exchange like gold. Since both of these options are more cumbersome than large denomination notes, storage and handling expenses will grow. This means the costs of running a criminal enterprise increases as does the odds of being apprehended.

Elimination of cash is a polarizing topic. For now I'm going to sidestep that debate because I think there's a more interesting topic to chew on: might a central bank be unsuccessful in its attempt to withdraw its own high denomination notes? Put differently, what happens if everyone just ignores a central banker's demands to retire the biggest bill?

Take the most popular high denomination banknote in the world, the US$100 bill. According to the Federal Reserve, there are 10.8 billion of these in circulation, or around $1 trillion in nominal value terms. Popular not only with criminals, the $100 bill circulates in many dollarized or semi-dollarized nations as a legitimate means of exchange in the absence of decent local alternatives. Say that the Federal Reserves wants to hobble criminals by cancelling all 10.8 billion notes. It announces that everyone holding $100s has until January 1, 2018 to trade them in for two $50s (or five $20s). After that date any $100 notes that remain in circulation will no longer be considered money. Specifically, they will cease to be recognized by the Fed as a liability.

Will this demonetization work? Consider what happens if everyone simply ignores the proclamation and continues to use $100s in trade. Say that by the January 1, 2018 expiry date, only $300 billion of the $1 trillion in $100 bills in circulation have been tendered leaving the remaining $700 billion (or 7 billion individual notes) in peoples' pockets.

So much for hurting criminals by removing the $100, right? We'd say that the central bank's demonetization campaign has failed. But not so fast.

Even though 7 billion $100 bills remain in circulation, the nature of $100 bill will have changed. Prior to January 1, 2018, the Fed maintained a peg between the $100 bill and all other denomination ($50, $20, $10, $5, and $1). This peg was enforced by the Fed's promise to convert any quantity of $100 bills into lower denominated notes and vice versa. After the expiry date, the Fed will no longer include the $100 in these pegging arrangements.

In addition to maintaining a fixed rate between the various denomination, the Fed also promises to peg the value of a dollar to a slowly-declining bundle of consumer goods (put differently, it set a 3% 2% inflation target). It does so by injecting an appropriate quantity of new currency into the economy via open market purchases or withdrawing sufficient currency by selling assets. When the Fed demonetizes the $100 note, it ceases to include the $100 in its consumer goods peg. This means it will no longer dedicate any of its assets to protecting the purchasing power of the $100 bill.

Given this new setup, as demand for $100 bills varies their value will float relative to both Fed dollars and the slowly declining consumer good bundle. Like bitcoin, which also has a fixed supply, fluctuations in the purchasing power of the $100 could be quite volatile. One day the $100 might be worth $110, the next it could be worth just $90. So even if the Fed has failed in withdrawing the $100, it will still have succeeded in imposing purchasing power volatility on criminals and other users of the $100. Volatile assets make for unpleasant and costly media of exchange and criminals will not be happy with these change.

By forswearing the $100, the Fed also ceases to act as a guardian of the quality of its issue of $100 bills against counterfeiters. The abdication of this function is especially important given that the marginal cost of printing a decent knock off of the $100 is probably just a few cents. Absent the threat of imprisonment, entrepreneurs will swarm to duplicate the $100, spending counterfeits into circulation and steadily reducing the purchasing power of the $100. After a few years of constant counterfeiting the $100 bill won't be worth much more than a few cents or so; the marginal cost of paper, ink, and printing. This hyperinflation will bring the nominal value of the original 7 billion in notes to just $700 million, down from $1 trillion.
Highest denomination note in various counties, sorted by US$ equivalent

Incidentally, we know this is a likely situation because of what has happened in Somalia. When Somalia's central bank was dismantled in the early 1990s, Somali shillings continued to circulate (see my blog post here). Over the next few years, warlords issued their own counterfeits which eventually drove the value of the shilling down to the cost of paper and transportation. William Luther has described this process here.

Along the way to hitting a terminal value of just a few cents, the $100 will lose any advantage it had previously enjoyed in terms of storage costs and handling. The moment a $100 falls to $49, the Fed's own $50 note becomes a cheaper note for criminals to use. And as the hyperinflation continues and the $100 falls to $19, the Fed's own $20 note will become preferred. The upshot is this: even if the Fed's January 1, 2018 expiry date fails to attract any $100s for redemption, competitive counterfeiting means that the $100 will inevitably cease to be used as the criminal economy's preferred medium of exchange.

Since criminals are rational and can anticipate that this sort of hyperinflation will ensue, they are more likely to tender their notes for cancellation prior to the original January 1, 2018 deadline. Better to get full restitution rather than lose all one's wealth.

Could criminals somehow police against hyperinflation by rejecting counterfeits? Militating against this would be the constant degradation of the note issue's quality due to normal passage of paper from hand to hand. In normal times, the Fed works behind the scenes to keep its note issue up to snuff, replacing worn out specimens with fresh new greenbacks. Once the Fed abdicates this role, $100 bills will quickly start to deteriorate. Picking the counterfeits out from a stack of bills will become more difficult, only making the job of counterfeiting easier.

In the face of this deterioration the mass of $100 bills may begin to fragment and lose fungibility. Fungibility is the idea that all members of a population are perfect substitutes. Well-preserved $100s that are easily identifiable as non-counterfeits may pass at a higher value than a slightly worn out $100, with well-worn and less identifiable $100 bills trading at an even larger discount. Without fungibility, it becomes far more difficult for a medium of exchange to do its job. Where a transaction with fungible $100 notes might be consummated in a few moments, it may take hours to grade a small stack of heterogeneous bills. The costs arising from non-fungibility may be so high that criminals will prefer to use relatively bulky $50 bills which, though possessing higher storage costs, will not be plagued by the requirement that each note be closely analyzed for quality.

So in the end, even if criminals ignore a central bank's deadline to tender notes for cancellation, they will eventually cease using the highest denomination notes through a more roundabout route. A central bank's renouncement of both its role as enforcer of the largest denomination's peg to other notes as well as its commitment to tend to that note's quality will set off forces that drive the purchasing power of those notes down to the cost of paper and ink, at which point they will be as good as demonetized.

Having settled whether a central bank can demonetize its highest value note, should it? That's an entirely different post. Or we can hash it out in the comments.

PS. An alternative story to a Somali-style hyperinflation is an Iraqi-style deflation. See Tony Yates on Twitter. I've written about the odd case of the Iraqi Swiss dinar here. How likely is an Iraqi scenario? Criminals would have to assume that a future monetary authority, maybe even the Fed itself, reverts its decision and undertakes to adopt orphan $100 bills as a liability at a price consistent with their previous purchasing power. This would give $100 bills a fixed value in the present.

PPS. On the topic of altering the relationship between high denomination notes and other notes, see my posts on high value note embargoes


  1. Yep. Good post.

    Typo: "(put differently, it set a 3% inflation target)" should be 2%?

  2. The Somalia story still puzzles me. I understand why the value of the notes didn't greatly exceed printing cost - what I don't understand is why counterfeiting was even possible. Gold coins can be melted down, but bank notes are useless for anything except money; Somalia was not on a "paper standard".

    1. Well, I suppose the next best use for banknotes once they have lost their role as money would be kindling.

  3. It was an interesting read, but my takeway is that "all is good".
    After all, no warlord in the world can print more bitcoins ;)

  4. Persuasive argued- an enthralling read.
    What would happen if a non-US entity decided to use the 100 dollar bill as its reserve asset? Suppose it is a feared and dreaded criminal operation which can ruthlessly deal with counterfeiters and defaulters and, moreover, that its Credit Creation is prudent. Are there any circumstances where this 'Mafia-Dollar'might hold its own?
    Suppose we move to a Bernie Sanders type global regime where the big monetary authorities agree to crack-down on tax avoidance by the 'super-rich'. In that case, invoicing in regular dollars creates a tax liability and becomes subject to capital controls (for example repatriation of profits to Sanctioned regimes may be disallowed) whereas the same transaction intermediated by the 'Mafia-Bank' in 'Mafia-dollars'may be less costly. Moreover, there is the risk that the 'Mafia-Bank' becomes an Enforcement Agency for Contracts and that the Judicial system gets hollowed out and disintermediated.
    I suppose, American global power is sufficient to combat such a threat to the Fed. However, it may be, the ECB would face a tougher struggle because of political turmoil on its borders.

    1. Your criminal operation might be able keep the $100 from hyperinflating due to counterfeiting, at least for a while.

      I'm wondering how it would deal with the fungibility problem. It would have to be able to fix worn out notes by withdrawing them from circulation and replacing them with new ones. If it has the ability to create decent substitutes, it might be tempted to spend new ones into existence, setting off the very inflation it was supposed to protect against.

      The other issue is whether the crime organization would dedicate any assets to protecting the value of the $100 from a sudden run. Central banks provide this sort of backstop. Without this feature, it will be far less risky for people to hold two $50 Federal Reserve notes than one $100 'Mafia dollar.' So in that sense, I think the Mafia Dollar would have troubles holding its own given the unlikelihood of a sufficiently large and credible backstop being deployed.

    2. I suppose I was thinking of 'hawala' operators and currency changers operating across dirigiste borders. At the time, Niki Kaldor's warning to the Indian Govt viz a crazy tax regime and the ban on gold imports creating a huge 'Black economy' chimed with his notion that money might be endogenous- i.e. hawala deposits might be giving rise to credit creation and this might be feeding back into domestic asset markets through off shore entities in places like Mauritius.
      A Mafia Bank prudent and professional enough to resist the temptation you describe could issue its own scrip and thus a demonetization event would merely fuel its growth as a one off event. It would not be the case, for the reasons you highlight, that such demonetized notes would have a permanent second life.

  5. Hey JP, I know this is the come to place for all things money, including history. At least that's how I think of it. I read this and thought it might interest you.

  6. I was surprised by the $ 1 trillion number

    Looked it up, and you are right of course

    Check out the considerable disparity in 20 year 'value growth' rates of the various denominations here:

    Of course, the $ 100 bill has a lot of leverage in translating volume growth to value growth. But still ....

    Any comment on that?

    1. That referred to the first graph

      The second graph also appears to show a volume growth rate that is probably out of whack with NGDP growth

    2. If I had to guess why use of the $100 has grown so fast I'd attribute it to dollarization and crime. In my experience, semi-dollarized countries tend to use the local currency for small denominations and $100s for large denominations.

  7. An analogous thing happens regularly (ca. every 20 years here in Switzerland) when CBs issue new notes. There is a transition priod during which old notes can be exchanged for new ones at member banks or the CB, after which old are officially obsolete. I've never seen anyone use old notes in commercial transactions.

    Does anyone use Francs, Marks or Lire?

    The first question you would need to ask is: why would anyone, let alone members of organised crime units, risk ignoring a change in legal denominations in face of near certain 99.99% devaluation over night? I just don't see it happening.

    I'd argue that everyone, including criminals as well as complete dollarised nations, just converts to the remainig legal bills during the transition period (up to 10 years for the euro) and future costs just get passed on to customers.

    1. I suppose what could happen is that banks / central banks in dollarised nations continued to accept $100 notes in payment after such a change while deploying its own assests to defend the exchange rate between the obsolete notes and the remaining official USD (or not). $100 notes would effectively become a new, local currency. But I don't see such a thing happening without the 'help' of a bank / banking system.

    2. I guess my point is, demonetisation is more of an orchestrated death during the official transition period, than a gradual inflation process as you describe it.

      And moving on to your actual question: should the US give up $100 bills? I guess the logistics of such an exercise could be set up as a global stimulus package albeit with possibly large, environmental costs.

      Pardon the chopped-up comment.

    3. Oliver, I agree with you that it would be odd for criminals to ignore a demonetization. But I think its interesting to ask what happens if they did ignore it. Or alternatively, what if the Federal Reserve gives an unrealistic redemption period; say just 24 hours to redeem? Would all the $100s that fail fail to make it to the redemption window in time be valueless or would the continue to circulate? (See my post on Swiss dinars mentioned above).

      You bring up an interesting point; maybe a central bank of a dollarized nation might choose to defend the peg between $100s and the two $50s. It would have to have an incredible amount of firepower given that it would be vouching for over $1 trillion worth of $100 notes.

    4. Wikipedia on the Italian Lira:

      All lira banknotes in use immediately before the introduction of the euro, as all post-World War II coins, were exchanged by the Bank of Italy up to 6 December 2011. Originally Italy's central bank pledged to redeem Italian coins and banknotes until 29 February 2012, but this was brought forward to 6 December 2011.

      Italy's Constitutional Court has now declared the law that shortening the exchange period of Italian Lira unlawful. Currently, Analyzes were launched by Banca d'Italia with the Ministry of Economy and Finance to define the procedures for the conversion of Italian Lira into Euro. Announcement regarding the detail information will be published shortly

      I doubt Lire are currently being used in any commercial transactions nor between 6.12.11 and 29.2.12.

      I guess I doubt the mechanisms you site by which a currency, money or notes might continue to circulate despite being officially demonetised really exist. So sure, you can ask 'what if', but reality informs me that it will always remain hypothetical question. What's the actual mechanism? What's the sine qua non of moneyness?

    5. Sorry, only now read the post on Somalia. Very interesting!

    6. Are there still banks in Somalia? Were there ever any? I figure money is older than the nation state and certainly older than central banks. So any truly general theory will have to make do without specifically mentioning either. Maybe warlords can be seen as a type of bank? Quite effective at collecting debts, I'd imagine.

  8. JP, this is a solution in search of a problem. Why convert 25% of the entire monetary base; and 75% of the value of currency outstanding? What is the social cost of the criminality, exactly? What is the value of social benefits?

    What is the value of eliminating the $100, as opposed to, say prosecuting fraudulent mortgage lenders? Or prosecuting HSBC's money laundering? Or shutting down the poppy fields of Afghanistan? Perhaps these actions would better benefit the fight against criminals.

    You've simply not established that there is a problem with criminals using $100 bills; and that withdrawing this bill would be the best solution.

    The answer is, simply, central banks can do whatever they like with the structure of their liabilities and assets.

    Still, you seem to be very keen to pick on the "little guy's" use of cash. There is one trillion in $100s outstanding -- how much criminal activity could there possibly be? What is the multiplier here?

    Answer wrong, and you just might be put on the "no-buy" list yourself. (You see how well the no-fly list works.)

    1. "You've simply not established that there is a problem with criminals using $100 bills; and that withdrawing this bill would be the best solution. " ... and ... "Still, you seem to be very keen to pick on the "little guy's" use of cash."

      Woah, at the outset of my post I explained that I would not be tackling whether large denomination notes *should* be withdrawn, only if they *could* be withdrawn.

      "The answer is, simply, central banks can do whatever they like with the structure of their liabilities and assets. "

      Then how do you explain the Swiss dinar?

    2. Hi JP. Nobody believes that this would end with the $100 note. If you accept the frankly unsupported assertion that cash=criminal, then we may as already start having "no-buy" lists. I'd suggest that cashlessness is too powerful a weapon to be given to authorities. If you yourself, JP, were put on a no-buy list (whether by mistake, injustice, or vendetta), who would speak for you?

      Re Dinar, it matters whether the CB is a private or public institution. Legal tender is a powerful and sticky phenomenon. Criminals could use Monopoly(tm) money if they like. It's a privately produced currency, and the Brothers Parker could enjoy the seignorage thereof, and the authorities could mandate ID checks and waiting lists to buy the game in response -- the moment Parker Brothers decide to retire to the Jersey Islands, it's over.

      Not so for public institutions. Sovereign states are infinitely-lived, and their obligations reflect such longevity. There's a cottage industry in buying up long-dead claims on sovereigns, and going to court to redeem. Dinar or $100, your counterparty matters.

      So, is the Fed public or private? There's your answer.

  9. Hey JP!

    You know that in the UK all notes are redeemable in perpetuity?
    So effectively they can't be withdrawn.

    Not sure, but I think that might have been so for US notes prior to 1950.

    I guess the point is that withdrawal of high denomination notes *could* only be done successfully with international co-ordination. I guess that'd be tricky to achieve.

    This whole ban cash thing is really gonna light up over the next few years. I think its difficult to predict its effects because its so deeply rooted. It'll be interesting to see what happens with the high denomination stuff cos that seems to be a soft target for the anti-cash folks.


    1. Hey Jon, thanks for the link.

      Different countries seem to have different policies over legacy note conversion. You may find the following interesting, which shows how and when various Euro nations plan to retire their legacy issues. Some have already done so, others never will.

  10. Last year the USA increased it's $100 banknotes in circulation by roughly 700 million. Conservatively 400 million of that increase went overseas. So while technically banknotes are a liability, they are never going to be "turned in". So in practical terms that is a $40 billion profit. While that may seem small compared to the trade balance of-$531.5 billion last year, it is at least in the same ballpark as the -$58 billion trade imbalance with Mexico.

    I think export of $100 banknotes exceeds every trade sector in the US economy with the possible exception of aerospace industry .

  11. I think you should include an analysis of the one country that is managing to withdraw it's HDN, Sweden. Sweden withdrew it's 100,000SEK banknote in the early 1990's (worth US$1800 at the time). Since the number of 1000SEK peaked in 2001 at 48.4 million they now have just over a million of the new banknote designs, and about 2.7 million of the old banknote design which will be invalid on Jun 30, 2016.

    There is a long list of things that Sweden can do, that the US government cannot legally do. One of the simplest is that they can refuse to publish their production quantities. Nobody knows for sure if Riksbank has produced 2-4 million banknotes of denomination 1000SEK, or if they have 30-50 million in their vaults which they can release if the banking system goes into crisis. In the USA the government cannot keep facts like that secret, as they are responsible under FOIA.

    In 2013 Iceland National Bank revealed that they had ordered 4 million of the 10,000ISK denomination (=40 billion ISK). That is a lot because the government was circulating only about 42.5 billion ISK in all their banknote denomination at the end of 2012. As of Jan 2016 almost half of those banknotes have gone in circulation, but smaller denomination notes have been reduced so that the total 50.7 billion ISK.

    1. Someone on twitter game me some interesting details on how Sweden has succeeded in reducing cash in circulation:

  12. A third and final comment on the Singapore Dollar.
    At today's rate 10,000SGD = 6,392EUR = $7116.
    While it is a massive value banknotes, it is important to note that only about 200,000 are circulating. A banknote for the 1% of the 1%.

    But Singapore has surprisingly high circulation of the 1000SGD banknote. At the end of 2014 they were circulating 13.742 million of the 1000SGD for a country of 5.5 million people (about 2.5 banknotes per capita). Switzerland, of course, dwarfs that number with 5 1000CHF banknotes per capita (which is a more valuable banknote).

    The Harvard report quotes the BIS summary in 2014 for the value of currency relative to GDP. But it is worth repeating:
    20.07% Japan
    15.67% Hong Kong SAR
    12.39% Russia
    11.55% India
    10.99% Switzerland
    10.33% Euro area
    8.82% Singapore
    7.74% United States
    6.46% Saudi Arabia
    6.19% Mexico
    5.04% Korea
    4.96% Turkey
    4.41% Australia
    4.01% Brazil
    3.80% Canada
    3.62% United Kingdom
    3.56% South Africa
    2.12% Sweden