I waved a $20 note in front of my macro class this morning. I said: "Is this a liability of the Bank of Canada?". They all answered "Yes".
That's what we teach them. We draw a balance sheet for the Bank of Canada. We put the government bonds it owns on the asset side, and the currency it has issued on the liability side. That's what the textbooks say. But is it right?
I think it's wrong. Maybe I should stop teaching that stuff.
Because it's made out of paper, and is more valuable than the paper it's written on, it certainly looks like debt. But that doesn't mean it is debt.
Let me tell you a counterfactual alternate history of money.
Once upon a time, in the olden days, people used gold as money. Gold was valuable because it was costly to produce, which limited supply, and because there was a demand for gold to use in fixing teeth and making airbag switches. The demand for gold to use as money merely added to that industrial demand and made it even more valuable.
The government used gold as money too. People had to pay taxes in gold, and the government used that gold to buy things. Sometimes, if the government wanted to buy more stuff than it had collected in taxes, it would borrow gold.
Then someone invented better things to use for fixing teeth and making airbag switches, so the industrial demand for gold disappeared. But the demand for gold to use as money meant it was still valuable, though not quite as valuable as before.
Then the government nationalised all the gold mines, so it had a monopoly. The government set the price of gold equal to the Marginal Cost of producing gold, so the value of gold was unaffected by the government monopoly (the government was neither more nor less efficient at mining gold than the private firms it replaced).
Then the government decided to abandon marginal cost pricing. The government would increase or decrease gold production, relative to the P=MC quantity, for various public policy purposes. The government would sometimes even make gold sales negative if it wanted to, by buying back gold it had previously sold.
Then an alchemist discovered how to make gold from paper, very cheaply. "Phew!" said the government, "It's just as well we nationalised gold production and made it a government monopoly!" The only difference the alchemist's discovery made was to reduce the government's cost of producing money. And it meant the government had to be even more vigilant to protect its monopoly, because the profits from producing "counterfeit" gold were now higher.
Do we list the houses sold by a builder as a liability on the balance sheet of that builder? No. Do we list the gold sold by a gold-miner as a liability on the balance sheet of that gold-miner? No. So where in my alternate history did gold become a liability of the central bank? Just because currency is usually made of paper, and liabilities are usually made of paper too, doesn't mean that currency is a liability.
The only difference between the endpoint of my alternate history, and where we are now, is that we use paper (and electrons) for government issued money, not gold. But the physical form of central bank issued money is irrelevant (unless it's very heavy or otherwise inconvenient etc.). Modern money is stuff that is very cheap to produce relative to its value, and isn't useful for anything else.
The fact that some paper currency was historically redeemable on demand into gold, and so was historically a liability in that sense, doesn't mean that paper currency is still a liability today. Sometimes, history really is bunk. Now that we are here, it doesn't always matter how we got here.
Now some central bank money is redeemable on demand into something else. If the Bank of Canada pegged the Loonie to the US Dollar, and promised to redeem Loonies at par in US Dollars, and promised never to break that exchange rate, and if that promise were written into the constitution, then you could perhaps argue that it is useful to think of the $20 note in my pocket as a liability of the Bank of Canada.
And you could also argue that the current 2% inflation target has some similarities to the pegged exchange rate example. It is as if the Bank of Canada had made the Loonie indirectly convertible into the CPI basket of goods, at a crawling peg that depreciates on average at 2% per year, with a 1% to 3% tunnel, plus some base drift.
If you think of that commitment to 2% inflation target as written in stone, you could argue that that makes the $20 note a liability of the Bank of Canada. If we wanted to hold fewer Bank of Canada notes, the Bank of Canada would need to buy them back in order to keep its commitment of 2% inflation. But it's still not a liability in quite the normal sense. Here's why:
If you list all the assets of a firm, and subtract all its liabilities, the difference should equal its net worth. That just doesn't work for the Bank of Canada. Bank of Canada currency pays 0% nominal interest, (and minus 2% real interest). Its assets earn positive interest. The Bank of Canada earns a profit on the spread. The net worth of the Bank of Canada is the expected present value of those (seigniorage) profits. A $20 note would only be a $20 subtraction from its net worth if the demand to hold that note disappeared tomorrow, and the Bank needed to redeem it to keep inflation on target. If the demand for Bank of Canada notes will never fall, then the present value of a $20 liability at 0% interest that need never be paid off is $0.
If I borrowed $20 at 0% interest, and I knew that the lender would never ask me for the $20 back, would it really be a liability? The demand for currency typically rises over time, in part precisely because it depreciates at 2% per year, so people would need to buy 2% more notes every year from the Bank of Canada just to keep up with inflation and hold their real stock of currency constant.
What about government bonds? If a Canadian government bond is a promise to pay Bank of Canada money, and that money isn't really a liability, and the government owns the Bank of Canada anyway, is the bond really a liability? Is government debt really debt?
Well, not in the same sense that private debt is a debt. The government of Canada could print off a very large number of $100 bills, equal in value to the national debt, plus interest, and pay off (most of) the national debt at a stroke, if it wanted to. (This wouldn't work for foreign currency bonds, or indexed bonds). But it wouldn't want to, under any normal circumstances, because people wouldn't want to hold that much 0% interest currency, so this policy would be incompatible with the 2% inflation target. Plus, the government's reputation and ability to borrow in future would probably be severely harmed.
But could the government just rollover the debt forever, issuing more debt to pay the interest on the existing debt? That depends. If the demand for debt is always growing faster than the rate of interest, then it can. This is definitely true for currency, which typically pays negative real interest, as long as the government keeps a de facto monopoly on currency, and doesn't try to persuade people to hold "too much" currency (in real terms). It may be true for government bonds too, as long as the government doesn't try to persuade people to hold "too much" in bonds.
What about privately issued money, like my chequing account at the Bank of Montreal? That's different. The Bank of Montreal is obliged to redeem my BMO money in BoC money, at par, on demand. BMO dollars are a liability of BMO in the same way that BoC dollars would be a liability of BoC, if the BoC were obliged, by law, to redeem my BoC dollars for US dollars at par (or at any pegged exchange rate).
(It's that asymmetric redeemability which gives the BoC, rather than BMO, the power to set Canadian monetary policy. And if the BoC pegged the Loonie to the US dollar, then it would be giving the Fed the power to set Canadian monetary policy. And if the Fed pegged the US dollar to the Loonie, then it would be giving the BoC the power to set US monetary policy.)
But suppose the Bank of Montreal had a de facto monopoly over chequing accounts in Canada, and could pay 0% interest on deposits, while earning monopoly profits from interest on its assets minus the administrative costs of running our chequing accounts? If BMO can borrow at 0%, and if it never needs to redeem demand deposits because the demand for demand deposits keeps on growing, would those demand deposits really be a liability? BMO's net worth would be the present value of the interest on its assets minus the transactions costs of servicing its demand deposits.
It is only competition between commercial banks that makes their demand deposits a liability in the normal sense of the word.
Just because something looks like debt doesn't mean it is debt.
(This post was inspired by, and is partly a response to, JW Mason's interesting post, which is well worth reading, even if I disagree with some of it.)
Worthwhile Canadian Initiative 
Counting down to Too Much Fed in 5, 4, 3….
Determinant: what surprised me, as I was writing it, is that my position is somewhat closer to (what I think is) the MMT position than I thought it would be. 5,4,3…
They used gold coins in the “good” old times not because the coins were valuable or pecious but because that made them difficult to counterfeit.
shorter Nick: My imaginary history of money imagines money to have been an asset for most of it’s history therefore money should still be thought of as an asset with now corresponding liability.
Nick,
Let’s suppose I’m the B of C’s magical accountant, I wave my wand, and a $20 bill appears.
I look at it and say “I now have an asset worth $20.” Is this right? It seems right to me, I can use that $20 to buy stuff.
Then I say “I need to record it on my books.” Do we still agree?
Being an accountant, I say “Everything has to be a double-entry, so to counterbalance my $20 asset, I have to have $20 somewhere else. There are only two possibilities, either liabilities go up by $20 or revenues go up by $20.” I wouldn’t put it past you to argue that the entire system of double-entry bookkeeping is incorrect, but I’m not sure that this is what you’re actually saying here, is it?
I can see the $20 being revenue – and then in the long term being added to the B of C’s accumulated surplus or owner’s equity. But have no idea why this might matter.
“If we wanted to hold fewer Bank of Canada notes, the Bank of Canada would need to buy them back in order to keep its commitment of 2% inflation.”
has exactly zero contradiction with
“If the demand for Bank of Canada notes will never fall, then the present value of a $20 liability at 0% interest that need never be paid off is $0.”
The latter just has stronger assumptions!
The philosophical question of “if it’s never going to be asked to be repaid, is it still a liability?” seems like a conundrum that applies to any very-long-lived institution that can make financial commitments, not merely central banks, actually. Worth comparing to public companies which never pay dividends, perhaps – are their shares liabilities? I should note that central banks appear to roll-over aggregate commitments rather than actually evade individual commitments.
Funny thing, after posting my last comment I went onto Youtube and this was the first thing recommended to me:
David Mitchell on the recession
It appears that he has a fine and subtle grasp of monetary policy.
Interesting thought experiment. I think that currency is indeed a liability of the central bank, but is more akin to deferred revenue than to debt.
When a private company receives money, but will (or may) have to perform some work in the future in order to earn that money, a liability is registered on the balance sheet. The Bank of Canada may, at some point in the future, exchange its government bonds to buy back outstanding currency in order to keep CPI (or NGDP, should its mandate change) from growing too quickly. Just because the currency is not redeemable on demand does not mean it is not redeemable at all.
Aha, it seems that balance sheet accounting does typically separate liabilities to shareholders into a separate column, so that one has assets = liabilities + equity, with equity being recognized as a funny kind of liability that isn’t really like other liabilities.
This is for public businesses, mind you, not central banks, but I don’t see why the intuition doesn’t generalize.
Didn’t Steve Randy Waldman (aka interfluidity) argue that state-issued fiat currency and state-issued debt are both better understood as equity?
@david
Public or private, long lived or short lived, shares aren’t liabilities.
“But it wouldn’t want to, under any normal circumstances, because people wouldn’t want to hold that much 0% interest currency, so this policy would be incompatible with the 2% inflation target.”
Probably, but it would be compatible with a -2% inflation target.
Gotcha. 🙂
Anton: sounds plausible to me. I wouldn’t know if something were gold or not.
Nathan: OK, so where exactly in my imaginary history did gold become a liability of the central bank?
Frances: “I wouldn’t put it past you to argue that the entire system of double-entry bookkeeping is incorrect, but I’m not sure that this is what you’re actually saying here, is it?”
I’m awfully tempted to answer “yes”!
Instead, the slightly more sober and careful me would answer “The BoC’s net worth has just gone up by $20 in nominal terms, and slightly less than that in real terms, depending on the elasticity of the money demand function.”
The David Mitchell thing was lovely. God, I’m so out of date with British comedy.
david: I tend to agree. Normally, for a private borrower, even if infinitely lived, the PV of the interest payments, discounted at market interest rates, equal the value of the liabilities. You can’t finance a permanent stream of consumption by borrowing. But if I could borrow at sub-market rates of interest, (which central banks can), then I could lend the proceeds and consume the spread. Private banks, unless constrained by competition, could do the same. So could any monopolist.
Neil: OK. Airmiles are a good analogy. Think of them as an interest-free loan. As long as the stock demand for unredeemed airmiles does not fall, the present value of redeeming them is zero. Even if the demand does fall to zero at some future date, the Present value of redeeming them is less than the face value of the liability.
Frances Woolley: “Everything has to be a double-entry, so to counterbalance my $20 asset, I have to have $20 somewhere else. There are only two possibilities, either liabilities go up by $20 or revenues go up by $20.”
There is a third possibility: it could be credited to equity…not exactly kosher by any means, but then again the BoC is not your usual firm…which I sometimes think is the point of MMT.
@Neil: “Just because the currency is not redeemable on demand does not mean it is not redeemable at all.”
Just because it is not redeemable does not mean that, if need be “prudent and wise men ” will abstain from doing what is needed to save the country. ( as said in the Report of the Bank of England on the 1825 Banking crisis, in preparation to the Bank Act of 1844). Everything is redeemable in at least something else.Unless,of course you forget that salvatio populi suprema lex esto…
david: “Didn’t Steve Randy Waldman (aka interfluidity) argue that state-issued fiat currency and state-issued debt are both better understood as equity?”
I think he did, and I think that’s right. I said something similar myself once. Maybe, for an inflation targeting country, they are more like preferred shares. We can always break the inflation target if we really need to.
Max: if you followed Milton Friedman’s Optimum Quantity of Money argument, and had a negative inflation target, so that nominal interest rates on bonds were pushed down to 0%, then seigniorage profits would go to zero, and currency really would be a liability of the central bank, because central banks would have zero net worth.
If the bank of candada is really committed to keeping the price level in canada rising 2%, then the paper money is a liability with a nominal interest rate of 0% and a real interest rate of -2%.
If the demand for currency should fall, then the bank of canada must pay off some of these loans.
If the demand for currency should fall to zero, then it must pay off all of those loans.
Rather than treating the paper money like gold, think the market basket of goods that defines the price level being like gold.
With a gold standard, the paper money is a debt payable in gold. Suppose rather than gold, you use gold and silver. Then gold, silver, and copper, then… a broad bundle of goods.
The paper money is still a debt. It is rather than a broad bundle of goods–the market basket of goods that defines the price level–replaces gold as medium of account.
Treating paper money like gold biases an economist to a base money quantity rule. If the quantity of base money is fixed, or grows at a fixed rate, then it is like paper gold and the government just spends the limited amount. Or, for that matter, if there is an unconstrained monetary authorty that just prints it and spends it, it is just like alchemy.
But if there is a commitment to a fixed price of something, then the paper money is a debt.
If you assume the demand for currency is always growing, for example, if you treat currency like it is the sole type of money, then maybe the debt framing isn’t very good.
But this is all wrongheaded.
Of course, I favor a strong commitment to nominal GDP, and oppose prohibition on private issue of banknotes. Put those two things together, and treating paper money like gold–an asset–is wrongheaded, and danger.
A general comment on money and government bonds: central banks like to pretend there is a firewall between base money and government borrowing. In this view, the bond market restrains borrowing by imposing a default risk premium…unless the central bank commits the ultimate sin of ‘monetizing’ the debt.
In reality, the bond market acts as if government bonds are backed by their respective central banks. The fact that the debt could be monetized means it doesn’t need to be. Hence a deflationary recession doesn’t cause a default risk premium to appear. Bond holders fear prosperity, not recession.
@Rowe: Wait, if you do think it is equity, then what inspired the post?
“Is this a liability of the Bank of Canada?”. They all answered “Yes”. And then I said: “No, it’s equity, and here’s why…?”
“if you followed Milton Friedman’s Optimum Quantity of Money argument, and had a negative inflation target, so that nominal interest rates on bonds were pushed down to 0%, then seigniorage profits would go to zero, and currency really would be a liability of the central bank, because central banks would have zero net worth.”
This is a slight digression, but I had a revelation recently that currency could be de-monetized. Suppose that currency were auctioned instead of being issued at par. Then it would be essentially a bearer bond, and could trade above face value. No more zero bound!
Paul: “There is a third possibility: it could be credited to equity”. Same thing.
Equity = the sum of revenues – expenses over all years.
Add the $20 to revenue and at the end of the fiscal year it gets added to equity.
Well, I was going to say you could call money degenerate debt. 😉
But then I thought, a debt is an obligation. Does the gov’t have an obligation wrt the money it issues? Well, yes. If I have some of its money, the gov’t is obliged to take it in payment of taxes, fines, and fees. If my tax bill is $600, they won’t accept a toilet seat. 😉
Sort of like MMT says, eh?
If a business is not in liquidation, we should account for it as a going concern, and we should value its products at cost until they are delivered to customers. The BoC is a business that produces liquidity services at zero cost. The future liquidity services of your $20 note have not yet been delivered, so we should account for them at cost, which is to say, for balance sheet purposes, we should treat them as if the value were zero. In other words, for liabilities currently on the BoC’s balance sheet, we should value them as if we lived in a world where their future liquidity services were worthless. In that world, the BoC would immediately need to shrink the money supply to zero in order to avoid hyperinflation and remain a going concern. It would have to sell all its assets to redeem its liabilities, and the value of the liabilities would thus be equal to the value of whatever assets it needed to sell to redeem them. Therefore, we should value $20 notes on the BoC’s balance sheet equal to the value of the assets they can buy. They really are liabilities, because their future services have not yet been delivered, and their value to the BoC as a going concern depends on their ability to yield those future services.
Nick, interesting post. It’s true that different liabilities (or promises) impose greater or lesser obligations on their issuers – there is a spectrum involved. For instance, some liabilities provide a first lien on assets, others are junior. Some require interest payments of their issuer while those like currency don’t. Some are payable upon demand, others are perpetual. Some liabilities come with sinking funds that guarantee some basement price, peg, or whatnot, while others are fully floating. Obviously an entity wants to issue those liabilities that impose upon them the least constraints: interest free, perpetual, floating, junior liabilities is the ideal. (sounds like stock to me)
Nick, you go to great extents to argue that the side of the liability-spectrum that imposes the least obligations on issuers actually imposes no obligation whatsoever, and therefore should not be labeled as a promise/liability. I’d argue that while they admittedly impose less-binding obligations than your average liability, financial instruments such as paper currency are still on the spectrum of promises/liabilities. The moment some financial instrument loses its character as liability and promises nothing is the point at which it hits 0.
It seems very important to a lot of economists that money be some intrinsically valueless token with no non-monetary uses, including being a liability/promise. Why is that? What wrench does money-as-a-liability throw into the works?
I wonder if there’s a political economy explanation. To enter the $20 as equity, it would first have to be shown as revenue for the Bank of Canada. Does the government really want to show the B of C making such large profits?
Base money isn’t really a liability of the central bank, but good monetary policy usually involves pretending that it is, e.g. by instituting a nominal GDP or inflation target.
Is there an analogy with monopolies acting competitive, in order to avoid competition?
Goven’ts act as if their currencies are liabilities, in order to avoid the real liability of using someone else’s currency?
Willem Buiter: “These monetary (base money) ‘liabilities’ of the central bank are not in any meaningful sense liabilities, because they are irredeemable…..” See:
http://blogs.ft.com/maverecon/2008/03/wanted-tough-love-from-the-central-bank/
@JP, economists want valueless tokens for currency because that’s one less superfluous reason not to trade, and economists are all about trading.
The ideal is to have synchronous trading, without the search costs. Those search costs are the reason we decouple things by using asynchronous trades, where money is the buffer. If people start valuing the buffer, they start holding the buffer, which can decrease throughput, leading to recession. It’s a pathology of the system that doesn’t happen with synchronous trading, and could theoretically be avoided.
jam: “Goven’ts act as if their currencies are liabilities, in order to avoid the real liability of using someone else’s currency?”
Spanish Dollars (pieces of eight) circulated freely throughout the American colonies and remained in use well into the 19th century. No problem. 🙂
An increase in demand for base money is like lending to the central bank, and that’s why base money is like a liability. Good monetary policy has the central bank acting like a financial intermediary; it should respond to an increase in demand for its liabilities by increasing its assets. A byproduct of this financial intermediation is an increase in the supply of base money to offset any increase in the demand, and the net effect on aggregate nominal expenditure should be zero.
Base money is only a ‘liability’ in the very broadest sense of the term. I don’t move in circles that use the term frequently, and so perhaps I’m defying convention when saying base money is not really a liability. Whatever. No need to quibble over words. The point is that base money never matures and can never be returned for redemption. Central banks might pay off debts by selling assets, but there are no natural market forces to discipline them otherwise. They are removed from ordinary market feedback by design. To get them to behave as though base money were a liability (in the narrow sense I use the term here), it is necessary to create enforceable rules for central banks to follow, e.g. an inflation target.
moving away from the snide (it was that kind of moment when i first read it), a couple points. First, Gold is an asset with no corresponding liability because it’s exists. It’s that simple. It has nothing to do with being money. When money is issued by a central bank, no matter what substance it is made of, it is a liability of the central bank. Not the underlying gold. If you melted down the gold, the liability of the central bank shrinks even though the physical asset exists just like when you rip up a dollar bill. It is true that in our modern world (to quote Michael Hudson) central bank money is a “du jure not a de facto liability”, but is still a liability. Note that this is true even if we acknowledge that the above history is a fantasy or if we pretend it is the truth.
Nick:
Compare two money-issuing banks: One stands ready to use its assets to buy back its money at par, and if the bank ever shuts down, people who hold its money will have a senior claim to the bank’s assets.
The other bank dumps all its assets in the ocean.
Which bank’s money will be valuable, and which will be worthless?
Reserve Bank of India on “I promise to pay ….” – (I think even the Bank of England does not answer it well)
What is the meaning of “I promise to pay” clause?
… The obligation on the part of the Bank is to exchange a banknote for coins of an equivalent amount.
So the central bank’s liability is to deliver the State’s liabilities 😉
Nice joke about the airbag switches by the way.
“The demand for gold to use as money merely added to that industrial demand and made it even more valuable.”
Corollory – which resulted of course in a social utility loss. (Mostly clearly seen during gold rushes.)
Nick,
are you really writing about Greece here?
Bill: suppose I sell a diamond. That diamond is not a liability to me. Suppose I sell the diamond, and also give the buyer a put option to sell it back to me at the same price any time in future. That put option is my liability. But the value of that put option is less than the value of the diamond, because it might not be exercised. Now suppose the strike price of the put option falls at 2% per year. The value of the put option is even less.
(Finance guys: what do you call that put option? American? European? Do there exist put options with strike prices that fall over time? Am I using words right here?)
Max @7.10: I basically agree. I like the bit about bond markets fearing prosperity, not recession. (The Eurozone is of course the exception that proves the rule, because they don’t borrow in their own money.)
david: It was JW Mason’s post that inspired mine. If money is like equity, it’s one that doesn’t pay dividends, and may (not must) be re-purchased in share buybacks, but at a falling price over time. It’s closer to equity than debt, but still a strange form of equity.
Min: the government, like my supermarket, and my creditors, is happy to take payment in BMO dollars. I’m not even sure if I can pay my taxes in BoC cash. Plus, my taxes are (roughly) indexed for inflation, so you could argue I’m paying in real goods.
Andy: you lost me there. Sorry.
JP: On reflection, I think the “put option” view of the liability that I sketched above is the way to go. I need to work on that perspective. I think I need a little help with it though.
“It seems very important to a lot of economists that money be some intrinsically valueless token with no non-monetary uses, including being a liability/promise. Why is that? What wrench does money-as-a-liability throw into the works?”
Good question. No obvious clear answer springs to the front of my mind. Maybe because it’s early 😉
Frances: Dunno. The way it’s done now is that the BoC only records profits as it earns interest on the assets it bought with the $20 note.
Lee: “Base money isn’t really a liability of the central bank, but good monetary policy usually involves pretending that it is, e.g. by instituting a nominal GDP or inflation target.”
I like that. Yes, that works for me. I need to bring the put option idea together with that.
jam: “Is there an analogy with monopolies acting competitive, in order to avoid competition?
Goven’ts act as if their currencies are liabilities, in order to avoid the real liability of using someone else’s currency?”
Maybe. But a government has to really screw up, Zimbabwean-style, to destroy even a crappy currency and letting it get driven out by a competitor.
Ralph: Yep, my views are close to Willem Buiter’s on this.
jam: yes on buffer stocks and asynchronous trading. Where (what literature) are you coming from with this perspective, by the way? Anything in particular? Or just a whole line of monetary economics plus your own thoughts?
Frances: I wouldn’t put it past you to argue that the entire system of double-entry bookkeeping is incorrect, but I’m not sure that this is what you’re actually saying here, is it?
Nick: I’m awfully tempted to answer “yes”!
Go for it, Nick! The US dollar has not been a liability since Nixon broke the link.
Nick: “Frances: Dunno. The way it’s done now is that the BoC only records profits as it earns interest on the assets it bought with the $20 note.”
Yup, and the seignorage revenue is just buried in general government revenue under ‘other’ so it’s really hard to actually work out how much it is. If it was recorded as revenue of the B of C it might be easier to work out. Or the B of C might be tempted to take the money and use it for something other than buying gov’t bonds. The more I think about it, the more it seems that politics rather than economics underlies the decision about which entity the revenue from printing money is attributed to.
Nick:
That’s not how a put option works.
The put option writer derives its income from the premium the buyers pays(“time value”). For the buyer, the payoff is the max(K-S,0), where K is the strike and S is the spot. It’s a kind of insurance policy.
You cannot say “the value of that put option is less than the value of the diamond” because you cannot compare option payoff and the spot price(see above).
“K” cannot fall by %2 as it is fixed by the contract.
Re. base money. When the central bank manufactures base money, it in fact writes receipts on debt paper (usually). So, the receipts possess value inasmuch as the debt paper may. That’s why “they” occasionally say that fiat money is backed by debt (debt paper is the CB asset, the receipts are its liability).
Andy Harless has nailed it, IMO.
(Finance guys: what do you call that put option? American? European? Do there exist put options with strike prices that fall over time? Am I using words right here?)
Your option is american by definition, because it can be exercised at any time. There exist many varieties of options with time-varying strike. I am not aware of any special name for an option with declining strikes, perhaps because the structure is not common. The closest match traded in practice might be a bermudan swaption with a declining strike schedule. However, “bermudan” refers to the exercise style rather than the strike (a bermudan option is exercisable on a set of discrete dates. Swaptions are often bermudan because the coupon dates make natural exercise dates. But your option is exercisable continuously, which is american.) Your wording seems fine.
Thanks Phil.
The post on this subject I ought to write next:
1. Lead with that comment from Lee Kelly above.
“Base money isn’t really a liability of the central bank, but good monetary policy usually involves pretending that it is, e.g. by instituting a nominal GDP or inflation target.”
2. Then say that inflation targeting is like selling an American put option, with a strike price declining at 2% per year, bundled in with the $20 note.
3. Then say that the $20 note is not a liability of the BoC, but the put option attached to it is a liability.
4. Then say something sensible and useful about the value of that put option. That’s the hard part for me, despite vjk’s attempt to help me above.
OK, let’s try this. The critical feature of a liability is not that it has to be repaid but that it has to be serviced. A perpetual bond would still be a liability. (After all, for a 100-year coupon bond with a substantial yield, the present value of the redemption proceeds is approximately zero, but we don’t consider it any less of a liability than a shorter maturity bond.) Money is like a perpetual bond that yields liquidity services. As long as the BoC is to continue operations, its money must continue to provide those liquidity services. Rather than servicing its debt with monetary payments, it services its debt with “in kind” payments in the form of liquidity services, and its business model requires that it behave in such a way as to continue providing those services. Of course, if the BoC decides it wants to suspend operations, it has the option of defaulting, and its creditors will recover nothing (i.e., there is no collateral and no recourse), but in so doing it will destroy its business model. In general, neither collateral nor recourse is considered a precondition for a debt’s being recognized as a liability. To the extent that we treat the BoC as a going concern, it has an obligation to service its liabilities.
Suppose I sell a car “as is”. That car is not a liability to me.
Suppose I sell a car, and write on the bill of sale “If for any reason you are dissatisfied, I will buy back this car no questions asked at the original price any time over the next month”. The car is not a liability to me, but the put option is.
It’s the same when the BoC sells a $20 note, except the put option lasts indefinitely, has a declining strike price, and has no force in law.
Andy: OK. I think I get it now. But what is the cost to the BoC of providing those “liquidity services”? Mainly: stamp out counterfeiting; don’t print too much new money.
I think I prefer my put option perspective.
Mike: If both monies were equally useful as a medium of exchange, and if there was no chance the demand for either would drop in future (or the supply would be increased), the ability of the issuer to use those assets to comply with the put option would be irrelevant.
Nathan: suppose I sold you 1 oz of gold, but at the same time promised to buy it back from you at the same price in future. Then there would be a liability attached to my promise. Same with paper.
Ramanan: there’s something similar in Canada. The Mint gets seigniorage from coins. The BoC gets seigniorage from notes. (I ignored coins in this post, to keep it simple). I’m not sure if the BoC is obliged to redeem in coins.
reason: the airbag switches was a deliberate anachronism, just in case somebody missed the “fake” history bit!
Yep, there are two social costs to using gold as money: it means less gold gets used in teeth; all the resources used to dig it up. Adam Smith said something similar IIRC, about how Britain could be better off if it replaced gold with paper and exported the gold.
This wasn’t really about Greece, though I think it’s relevant.
Arthurian: I should stay away from the temptation to say things too wild!
Frances: the BoC reports its profits on the annual accounts, IIRC. Last time I looked it was $2B annually. Probably more now.
A typical variable strike price option is the cliquet option.
Both swaptions(puttable or callable swaps) which are in 99% of cases fixed strike price options and cliquets are exotics.
If one assumes the vanilla put with varible stike, one has something other in mind than the vanilla put.
One is free of course, given a willing counterparty, to create any sort or crazy contracts (and does).
The traditional view of money as a liability of the central bank comes from thinking of the central bank as a regular bank. Just like how regular banks issue loans and invest in securities, which comprise the bank’s assets, the central bank issues loans (as lender of last resort) and buys government securities in open market operations. These are the assets of the central bank. The central bank has two sources of funds to finance its purchases of assets: reserves from private banks, and newly created money. Thus, as the source of financing, money is a liability of the central bank.
It is true that central banks no longer have to exchange money for gold on demand. But money is still redeemable in some sense–its just that it is redeemable for money, and the central can simply “buy” back your money by issuing more money. In that regard the central bank acts as the only legal ponzi scheme–it is exactly the same as a bank who pays out liabilities by borrowing even more.