When I was writing my last post, on commercial and central bank solvency , I wasn't sure whether it was worth posting, because maybe everybody already knew this stuff. But I decided to post it anyway. Now I'm glad I did.
I got some good comments from JKH, which opened up a new avenue to explore; John Palmer posted it at his blog EclectEcon , and opened up some new questions in a different direction; and now I have just read this article about the Bank of Canada's balance sheet from John Paul Koning in the Financial Post.
It's worth reading the FP article, because it does point out how much the Bank's balance sheet has changed, and how unprecedented this is. But it also repeats the theory that the value of the Bank of Canada's liabilities (largely paper money) depends on the value of the Bank of Canada's assets.
A liability is only as good as the asset that backs it up. The $20 you’re holding is backed by various assets held in the vaults of the Bank of Canada in Ottawa. In times past the asset side of the bank’s balance sheet had a large gold component. Over the years gold lost its popularity with central banks and was replaced by government bills and bonds. As late as August of this year, the bank held assets of $22-billion in government T-bills and $31-billion in long-term bonds to back the cash in Canadians’ wallets, under their beds and in their deposit boxes.
This has all changed. Over the last three months, the bank has sold off a large part of its government T-bill portfolio and replaced it with assets classified as “other.” In August this “other” category comprised a miniscule 0.4% of the bank’s total assets, or about $200-million. It has since ballooned in size to an impressive $32.4-billion. Last week “other” surpassed government bonds to become the largest component of the bank’s assets, about 42% of the total. At the same time the bank has sold off $11-billion worth of government T-bills, which now make up just 15% of the bank’s assets, down from a hefty 41%.
Why does this matter?
If the Royal Bank sold 42% of its performing loans and replaced them with mystery assets, it’s probable that a portion of depositors would get nervous, close their accounts, and go to the Bank of Montreal to do business.
Likewise, the Bank of Canada’s decision to lower the stability and transparency of its balance sheet may incite people to sell Bank of Canada liabilities. The result would be a drop in these liabilities’ value, which is just a different way of saying inflation, or a decline in the purchasing power of money.
This is a good statement of a theory I call the "backing" theory of money: the value of money is determined by the value of the assets which "back" that money.
To explain what is wrong with that theory, let us suppose that all of the Bank of Canada's assets go bad, all $50 billion become worth $0, so the Bank has no assets to back its $50 billion in paper money liability. This really is a worst case nightmare scenario. What would change?
The liabilities of the Royal Bank, the Bank of Montreal, and all the other commercial banks are redeemable, mostly for Bank of Canada paper money. That paper money, the so-called liability of the Bank of Canada, is not redeemable, and hasn't been for decades. If I go to the Bank of Montreal, clutching my cheque book, and demand my money back, they are required to give me Bank of Canada money. If I go to the Bank of Canada, clutching a $20 bill, and demand my money back, I think they would just call security (I'm scared to try it, since I might want to visit there again sometime). What's more, the Bank of Canada doesn't even promise to pay any interest on that paper money liability. So what sort of a liability is it? And how would I be at all affected if the Bank of Canada's assets suddenly became worthless?
Actually, there is one way I would be affected, but as a taxpayer, not as a holder of money.
Last year, the Bank of Canada paid about $2 billion to the government of Canada. This $2 billion represents the interest earned by the Bank of Canada on its assets. If all those assets went bad, that would be $2 billion less per year that the government would get from the Bank, so the government would have to raise taxes by $2 billion to compensate. Not nice, since that's about 0.15% of GDP, but it's not the end of the world.
The only way I could be affected as a holder of money is this: it is theoretically possible that the government would refuse to raise taxes, or cut spending, and insist that the Bank of Canada print an extra $2 billion of paper money per year to compensate the government for the lost income from the Bank of Canada. That extra $2 billion, divided by $50 billion, is 4%, and would mean that the money supply would now grow 4 percentage points faster than otherwise, so the inflation rate would be 6% rather than 2%.
Worthwhile Canadian Initiative 
Nick, nice to see an article I wrote spotlighted in a blog I enjoy reading.
Let’s say the BoC exchanges its $50 billion in bonds for $50 billion in bananas. A silly scenario, but worth considering. The next day the BoC’s assets have gone soft and are worth $30 billion, and the day after they’re rotting and worth $10 billion. By day four they’re worth $0. Now I can guarantee you that once this news gets out to the public, currency traders will dump their C-dollars as quickly as possible for other currencies. Canadians will sell their dollars as quickly as possible for stuff. Who wants a banana-backed currency?
As to your point on redeemability, C-dollars have limited redeemability because members of the Canadian Payments Association can engage in reverse purchase and resale with the BoC whereby they submit their dollars for bonds. These rarely happen and are at the initiative of the bank. Rare or not, none of the members engaging in reverse PRA’s would be happy to know they’d be getting rotten bananas rather than bonds. There’s a good chance they’ll sell c-dollars for US dollars since the Fed’s equivalent – reverse open market operations – at least return a real asset; bonds.
But the bigger issue is that liabilities don’t have to be redeemable on demand to reflect the value of the assets that back them. Perpetual bonds are non-redeemable liabilities and they reflect the value of the assets backing them. Stocks are the best example. You can’t bring your stock certificate to Nortel and ask to redeem them for x% of their assets, say a desk and some pencils. Yet Nortel stock, through the buying and selling on the market, reflects the value of its assets.
Even stocks have limited redeemability; this comes into effect when the company is unwound or goes bankrupt. Stocks, after all, are a claim to the firm’s remaining assets. Same with Bank of Canada notes. The Bank Act says that “the Bank has the sole right to issue notes and those notes shall be a first charge on the assets of the Bank.” So in the case of wind-up, note holders can indeed redeem their notes for BoC assets, and they get priority. Like stocks, BoC notes have “end of the road” redeemability, and that’s why assets are indeed important.
By the way, I completely agree with your analysis about the BoC paying less to the government in profits were assets to go bad. But to say that backing doesn’t matter means you have to defend a scenario in which the Canadian dollar doesn’t fall when Mark Carney announces to the world that Canadian dollars are now banana-backed, not bond-backed. If you can justify that one, I’d be convinced that backing is irrelevant.
John: Thanks. I’m pleased that someone like you is reading WCI, and I’m very pleased you have joined in the debate here.
I like your banana example. It’s only “silly” examples like that which are extreme enough and simple enough to force us to think clearly. I also like your analogy between money and stocks. Sometimes I call it the “backing” theory of money, and sometimes I call it the “equity” (in the sense of stocks) theory of money. (Fisher Black is one proponent of the “equity” theory of money, and he’s wrong!)
I’m going to duck what you said about PRAs for the minute, because my head is not as clear as it should be on that question.
First I’m going to throw three punches to try to soften you up. Then I’m going for the guts of the issue.
1. In normal times, the assets of the Bank of Canada are government of Canada bonds. I take it you are quite happy if the Bank of Canada notes are 100% backed by such bonds. But what are those bonds? They are promises to pay….Bank of Canada notes! So Bank of Canada notes are ultimately backed (in normal times) by promises to pay more Bank of Canada notes. If I believed the backing theory, I think I would feel more confident knowing the BoC had a lot of bananas in the basement. If I gave you a NR note, and told you it was 100% backed by bonds which promised to pay more NR notes (and even redeemable on demand for such bonds, how much would you pay (in real goods) for a NR note?
2. Take the example of Yap stones (like the one in the Bank of Canada’s garden — which I’m told is actually part of its forex reserves). No intrinsic usefulness whatsoever (except as money), and backed by absolutely nothing, yet they served quite well as money. Cowrie shells are another example. Backed by nothing, of very limited usefulness as jewelry (though that is probably more a Veblenesque reflection of their logically prior value as money than of their somewhat suggestive shape), yet they too served well as money (until the supply increased causing hyperinflation).
3. Suppose the Bank of Canada has been lying about its assets for the last 50 years. Suppose they actually don’t have any bonds in the basement at all, because various governors have been gambling them away on drunken sprees in the Caymans. All successive governors (and the government) have colluded to keep this scandal secret, and threatened the auditors with midnight visits from CSIS. How would we ever know? How would this ever matter to us, as long as it were kept secret? (Those were rhetorical questions.) Now there is something wrong, not just empirically wrong, but logically wrong, with a fundamentalist theory of money (“fundamentalist” in the Graham and Dodds sense, not the religious sense) in which mere belief that there are real assets backing the money is just as effective in keeping money valuable (not just now, but for the indefinite future) as the real physical truth of the existence of those assets.
4. Let’s compare three things: paper money with no backing; a Ponzi scheme; and a closed end mutual fund which is well-managed, pays no dividends, never does buy-backs, reinvests all earnings, and has a termination date in the indefinite future.
The mutual fund is 100% backed (by assumption). According to the backing/fundamentalist/equity/Graham-Dodds theory of value, shares in that mutual fund should be equal to the value of the underlying real assets, and should appreciate at the market rate of return (call it the rate of interest). Yet if the rate of interest exceeds the growth rate of the economy, you get a contradiction. The market cap of the mutual fund, as a ratio to GDP, would approach infinity as time went to infinity. So everyone would eventually dump their shares at some time in the future. Knowing that, nobody would buy the shares now, so the shares would be worthless, despite their backing.
Now look at the Ponzi scheme, which has 0% backing by assumption (Mr Ponzi consumed all the assets from the first $100 loan). But nobody knows this. If Ponzi pays a rate of interest higher than the growth rate of the economy, the market cap of the Ponzi scheme, as a ratio to GDP, approaches infinity as time goes to infinity. Net wealth of Mr Ponzi’s depositors becomes infinitely high relative to their income, so they want to withdraw their deposits to buy stuff. The Ponzi scheme collapses. But suppose, just for the sake of argument, that Ponzi can borrow at an interest rate below the growth rate of the economy. In this case, the market cap of the Ponzi scheme, as a ratio to GDP, falls over time. The Ponzi scheme never collapses.
Now look at paper money, which is unbacked (by assumption). Because holding money is useful, as a medium of exchange, people are willing to hold money even at a zero nominal rate of interest, and a negative real rate of interest (if inflation is positive). As long as the nominal growth rate of GDP is positive, the market cap of money, as a ratio to GDP, stays finite. It is stable.
Paper money is like a Ponzi scheme, except people are willing to hold it despite an interest rate which is lower than other assets, and lower than the growth rate of the economy, and they are willing to accept such a low interest rate because it is useful as a medium of exchange (a liquidity premium).
Sorry. I’ve rambled far too long.
I wasn’t aware that Fisher Black held those views, will have to read up on that.
Let me take up your points, or should I say punches, one by one.
1. Your point is that Bank of Canada notes are backed by government bonds, which are ultimately just promises to pay more notes, and therefore you’d just as rather have the banana backing. Let me argue why I think most people would prefer bond backing.
It’s best to isolate the ultimate backing of something by considering a scenario in which the issuing institution goes bankrupt or chooses to be unwound. Say the Federal government and the Bank of Canada decide to stop doing business. People bring all their dollars to the BoC to claim bonds, since notes are a first claim on all assets. With their bonds in hand people now go to Parliament. In our government unwinding scenario, bondholders are creditors, and like any creditor they have a claim to governments assets. Some bondholders might get title to land, some a building, a few might earn partial ownership to the rights for the Canadian flag etc.
Now if BoC notes are banana-backed, the only thing in our unwinding scenario note holders will get will be a couple of rotten bananas. I’d rather have ownership in a building or any other former government-owned asset than a rotten banana, and that’s why I want notes to be bond backed, not banana-backed.
2. Yap stones and cowrie shells are commodity money, and I’d guess they were probably useful or prized prior to becoming money. If a bank was created that backed its paper money with yap stones, the price of the paper money would come to be determined – in part – by the value of the backing.
3. Enron’s liabilities kept their value even though their assets were a mess. Same with Penn Square Bank in 1982, which lent billions to drillers in Oklahoma with bad credit and iffy collateral. Through shady accounting and likable personalities they manipulated their image and were able to fool the world, but at some point word got out and investors sold not just their shares but their bonds. Likewise the Bank of Canada could burn all its assets and maybe keep up the ruse up for a few years, but at some point the news would get out. It’s all about perceptions and the manipulation thereof.
4. I was a bit confused by point four but it seems to me that you are trying to show the irony of a backed fund collapsing while an unbacked ponzi scheme never collapses, and therefore backing is irrelevant. Do you mean to say that you think backing is unimportant for assets like stocks and funds? I thought you just objected to a “backing” theory for money. Do you object to “backing theory” being applied to Royal Bank deposits (or old fashioned Royal Bank notes) or just government money? At what point in the spectrum of assets does backing become irrelevant?
Your mutual fund collapses in the present because the people in your example seem to have perfect knowledge about the future. If they had perfect knowledge, how could one mutual fund manager consistently beat the market, or how could anyone be duped by a ponzi scheme?
I would agree with your comparison of paper money to a ponzi scheme, but only when this paper is unbacked. Both can only be kept going by deception. Well backed paper money is not a ponzi scheme.
Thanks John.
Don’t trust my memory on Fischer Black; I even spelled his name wrong!
1. If the BoC is wound up, and people exchange cash for bonds, and then take those bonds to the government for redemption, it is not obvious to me how the government would translate dollars (which is what the bonds promise to pay) into buildings etc. In a normal commercial bankruptcy, the assets are sold for cash, which is then divided up among the bondholders pro rata. But if the Canadian dollar no longer exists, it is hard to imagine how this could happen. If the Canadian government went bankrupt, I expect the buildings could be sold in US dollars, and then just divided pro rata. But if the Canadian government were not bankrupt, what would constitute full payment of the bondholders?
2. Cowrie shells probably had some value as jewelry before they becae used as money. But if the demand for shells as money was large relative to the demand for shells as jewelry, their value would have increased when they became used as money. If it doubled, for example, then you could say that cowrie shells had 50% backing. But even then, how can you explain the other 50% (the unbacked portion) of the value of cowrie shells? With the Yap stones, it is really hard to think of any conceivable use for those things, except as money.
3. When people learned that Enron shares and bonds had no backing, they learned that Enron was a Ponzi scheme, and an unsustainable Ponzi scheme. It was unsustainable because nobody will hold Enron shares/bonds except at a rate of interest which exceeds the growth rate of the economy (see below).
4. I wasn’t as clear as I should have been on point 4, but you got what I was trying to argue (the irony of a backed mutual fund collapsing etc.). Your question “At what point in the spectrum of assets does backing become irrelevant?” is the right one to ask. Let me try to answer it:
Think of the spectrum of assets very broadly: everything from stocks and bonds to cars, paintings, and CDs (I mean the ones that play music). Each asset has an equilibrium desired rate of pecuniary expected return (dividends plus capital gains) at which the existing stock of that asset will be just willingly held. Line up all those assets from the highest to the lowest rates of return. At one extreme we have risky illiquid stocks, which will only be held if people expect a very high rate of return. For these assets, backing certainly matters. I want to be sure there are some real assets to pay the dividends, or to be bought out in an eventual take-over. At the other extreme, my CD collection has assuredly negative capital gains, and pays no pecuniary dividend, but I like owning it despite the strongly negative rate of return (OK, it pays an in-kind dividend, of listening pleasure). Same for paintings, only they might or might not appreciate. My car will depreciate, and pays negative dividends (mechanics bills). We don’t ask what assets are backing CD’s, cars, and paintings. They are their own backing.
Money (paper currency) in Canada sits in the middle of that spectrum. It pays no dividend, and depreciates annually at 2% real (inflation), so it has an expected real rate of return of minus 2%. Yet I willingly hold it. One metaphor is to say that holding money yields services in kind, like a refrigerator, painting, or CD. Liquid assets have lower rates of return in equilibrium, and money is the most liquid of all assets. Even in Zimbabwe, where money must have the lowest rate of return of all assets in the spectrum, people are still willing to hold some of their wealth in the form of money. Backing is not needed for money, just as it is not needed for refrigerators.
Where exactly in the spectrum is the dividing line, between those assets which do and do not require backing? This is where Mr Ponzi comes back into the story. If Mr Ponzi can only borrow at 10% (so his liabilities have a desired equilibrium rate of return of 10%), and if he has no backing, so must roll over his debt, then his total debt must grow at 10%. This is faster than total GDP growth, so Ponzi debt is growing relative to GDP, which pushes up the required rate of return still further, and it is not sustainable (it’s a bubble, and must crash eventually because people are not willing to hold 100 years’ worth of income in Mr Ponzi’s liabilities). But if somehow, Mr Ponzi could borrow at (say) 1%, then his debt would grow at less than GDP, so it would be sustainable. If Mr Ponzi’s liabilities were the most liquid in the economy, he could borrow at a very low rate of interest, perhaps even negative. People would pay to lend to Mr Ponzi.
That’s exactly what the Bank of Canada does. People pay 2% real (minus the inflation target) for the privilege of lending to the Bank of Canada, so long as they get this nice convenient bond called a $20 note. People in the past have been willing to pay much higher amounts for the privilege of lending to central banks (only they don’t want to lend as much at such rates, so the real money supply is lower when inflation is higher).
The exact cut-off between backing and non-backing (unsustainable Ponzi and sustainable Ponzi) is where the rate of return equals the growth rate of GDP (I think).
It is ambiguity around that cutoff (which we have seen over the last few years when interest rates were very low, and about the same as GDP growth) which causes asset bubbles, and financial crises. But that’s another story, and my head’s not clear on it.
I see Garth Turner has taken up your article on his housing crash blog!
RBC Bank President Gordon Nixon – Salary $11.73 Million
$100,000 – MISTAKE (FISHERMEN’S LOAN)
I’m a commercial fisherman fighting the Royal Bank of Canada (RBC Bank) over a $100,000 loan mistake. I lost my home, fishing vessel and equipment. Help me fight this corporate bully by closing your RBC Bank account.
There was no monthly interest payment date or amount of interest payable per month on my loan agreement. Date of first installment payment (Principal + interest) is approximately 1 year from the signing of my contract.
Demand loan agreements signed by other fishermen around the same time disclosed monthly interest payment dates and interest amounts payable per month.The lending policy for fishermen did change at RBC from one payment (principal + interest) per year for fishing loans to principal paid yearly with interest paid monthly. This lending practice was in place when I approached RBC.
Only problem is the loans officer was a replacement who wasn’t familiar with these type of loans. She never informed me verbally or in writing about this new criteria.
Phone or e-mail:
RBC President, Gordon Nixon, Toronto (416)974-6415
RBC Vice President, Sales, Anne Lockie, Toronto (416)974-6821
RBC President, Atlantic Provinces, Greg Grice (902)421-8112 mail to:greg.grice@rbc.com
RBC Manager, Cape Breton/Eastern Nova Scotia, Jerry Rankin (902)567-8600
RBC Vice President, Atlantic Provinces, Brian Conway (902)491-4302 mail to:brian.conway@rbc.com
RBC Vice President, Halifax Region, Tammy Holland (902)421-8112 mail to:tammy.holland@rbc.com
RBC Senior Manager, Media & Public Relations, Beja Rodeck (416)974-5506 mail to:beja.rodeck@rbc.com
RBC Ombudsman, Wendy Knight, Toronto, Ontario 1-800-769-2542 mail to:ombudsman@rbc.com
Ombudsman for Banking Services & Investments, JoAnne Olafson, Toronto, 1-888-451-4519 mail to:ombudsman@obsi.ca
http://www.corporatebully.ca
http://www.youtube.com/CORPORATEBULLY
“Fighting the Royal Bank of Canada (RBC Bank) one customer at a time”
Hi Nick, not sure how far you want to go with our debate, but I can’t resist adding a bit more on points 1 and 4.
1. If the government winds up, each bondholder can get a proportional ownership interest in former government assets equivalent to their share of government bonds. This could be represented as a title to x% of all assets. This title could be kept by members of the public or sold for $US or whatever medium of exchange has sprung up to replace the loonie, or it could be bartered for stuff. There are probably many ways to facilitate this process of distributing government assets, but I think my point holds up that it’s better to own a bond-backed currency than a banana-backed currency.
4. Your argument is that assets are placed in a spectrum according to expected rate of return, and that there is a line somewhere in the spectrum, above which assets have backing (like illiquid stocks) and can be evaluated in such a manner, but below which backing is not important (like refrigerators). You described your cd collection as an example of the latter.
I would agree that backing is not important for cd’s, but not because their yield is below a certain rate. Let’s say you created redeemable paper claims to your cd collection which you distributed in the market. These paper claims, your liabilities, would earn their value according the price of the underlying cd’s – ie. they would have backing and evaluated with respect to that. But they would also fall in the same low yielding part of the spectrum that the actual cd’s would. That’s why I don’t think the return of an asset has anything to do with its ability to be valued by its backing.
I would say that the spectrum on which “backing” theory is relevant is like this: any thing which appears on the liability side of a balance sheet can be evaluated with “backing theory”, otherwise it cannot be evaluated in such a manner (like a fridge). Claims to things can be evaluated with backing, but not the things.
If you agree that stocks can be evaluated by looking at backing, how about bonds? The price of bonds depends on the ability of a firm’s assets to generate a decent enough return to pay interest and return principle. If you accept that bonds can be evaluated in that way, the circulating notes of a private bank (take the Royal Bank of Canada’s old 1850s issues) can also be evaluated with respect to backing since they are just another liability of the bank. If you accept that, then the BoC’s liabilities should also be evaluated in such a way since the only difference between BoC notes and the old RBC notes is redeemability and monopolization.
John: I don’t think either of us is going to convince the other, but our debate has nevertheless been very worthwhile to me. It has forced me to clarify my thinking.
One last stab: if the Royal Bank had a (partial) monopoly on note issue, then it could pay a less than market rate of interest on its notes, so there would be a value to their liquidity services at the margin (interest rate on notes + liquidity services at the margin = bond rate of interest). Ignoring administrative costs of the note issue, the Royal Bank would have monopoly profits equal to the spread between the interest on notes and the bond rate of interest. The notes are “backed”, if you like, by those monopoly profits, as well as by the bonds. If we assume the notes pay no interest, they are fully “backed” by the monopoly profits. This is then exactly like the Bank of Canada, except for one big difference: the Royal Bank notes were redeemable (I think). That means there are two equilibria: one where people hold the notes; the other where people expect a run and there is a run. (Diamond Dybvig model of bank runs). The Bank of Canada notes, being irredeemable are not subject to runs.
And all the above is true of we replace “notes” with “deposits”. In other words, if the BMO had a monopoly on checking accounts, and paid zero interest (and no admin cost), they could issue unbacked chequing accounts (provided the BoC stood ready to prevent runs, to rule out the second equilibrium).
Dr. Rowe,
I don’t have a background in economics, so I apologize for what are likely very basic questions, but here goes:
1) With Canadian notes NR, why does the BoC have any asset-backing at all (CND$ bonds or banannas)?
2) If the BoC issued notes without increasing it’s assets, is that what’s meant by “Printing more money” and is bad?
3) Using JP’s Bank Act quote, it would seem a CDN note entitles me to a claim on the BoC assets, but does anything anywhere legally extend that claim onto the Government of Canada’s assets? I assume they are different entities.
Please note, I accept your position and don’t need to be convinced of it — I’m looking instead for further education.
Alan: Good questions, despite your lack of economics background.
1. Suppose there were a fall in the demand for currency. The Bank of Canada would want to buy back some currency because if it didn’t there would be an increase in inflation. It can use the assets it owns to buy back the currency. But it has far more assets than could ever conceivably be required for this purpose. I think the only real answer is that it keeps the accountants happy. And it doesn’t really make any difference anyway, since the government owns the Bank of Canada, it really owns all the Bank of Canada’s assets, and gets the interest on those assets in either case. But by keeping the assets on the Bank’s books, rather than on the government’s books, does help symbolically to keep the Bank independent of government.
2. Maybe, but the terminology is inexact. Currency is always printed. If the Bank prints money and buys assets, that’s called an “open market operation”. If the Bank prints money and just gives it away, that’s called “helicopter money” (print money, load it into a helicopter, and throw it out the door). Is it bad? Depends. If you increase the supply of money more than the demand for money increases, it causes inflation. If you increase the supply less than the demand, you cause deflation. You have to print (or burn) just the right amount, and that depends on circumstances.
3. I have no idea!