One of the things I was a little disappointed about in the recent money and banking controversy was that nobody (IIRC) picked up on my (it's not original to me) point about asymmetric redeemability.
The Bank of Montreal promises to redeem its (demand deposit) dollars for Bank of Canada dollars, while the Bank of Canada does not promise to redeem its dollars for Bank of Montreal dollars. It's a one-way fixed exchange rate system between two competing currencies. I see that fact as absolutely crucial if we want to understand why the Bank of Canada controls Canadian monetary policy and the Bank of Montreal does not. Nothing else really matters.
Suppose, just suppose, that the US Fed decided to peg the US dollar to the Canadian dollar at par. And the Bank of Canada just shrugged its shoulders and made no such commitment the other way. Macroeconomists generally believe that the Bank of Canada would then control monetary policy both for Canada and for the US. If the Bank of Canada wanted to tighten monetary policy to fight inflation in Canada, that would be unfortunate for the US if the US were in recession.
1. If that were to happen, in what important ways would the US Fed be different from what the Bank of Montreal is now?
2. If that were to happen, in what important ways would the Bank of Canada's control over CanAm monetary policy be different from its current control over Canadian monetary policy?
3. In what important ways does the the institutional structure of the payments system matter for implementing monetary policy, if the US continued to use its own existing payments system despite pegging its dollar to the loonie?
(Perhaps they are all the same question.)
Nick
You won't agree, but the prof. who taught me most of what I know about finance (don't dismiss him because he is a finance guy! )uses an analogous example to argue that money as a medium of account is perhaps more relevant than money as a medium of exchange.
Point being that both the BoC/ BoM can create the medium of exchange, but only BoC defines the medium of account. I see this as a corollary to asymmetric redeemability, though you may not agree.
http://jrvarma.wordpress.com/2012/01/19/the-many-different-kinds-of-fixed-exchange-rate-regimes/
Posted by: Ritwik | April 03, 2012 at 03:26 PM
Nick: "the Bank of Canada does not promise to redeem its dollars for Bank of Montreal dollars"
Well it kind of does. For most intents and purposes, I think that's the meaning of deposit insurance.
To answer you first question, the BoC wouldn't guarantee the US dollar.
Posted by: K | April 03, 2012 at 03:39 PM
Ritwik: "Point being that both the BoC/ BoM can create the medium of exchange, but only BoC defines the medium of account."
That deserves one of my rare "Hmmmmm's". Even before reading that link.
K: OK. But that only really applies in a crisis, I think, when the Bank of Canada might temporarily lose control of monetary policy, and not be able to do what it wants to do, because it needs to bail out a bank?
Posted by: Nick Rowe | April 03, 2012 at 04:00 PM
Nick: I would say it applies only in all circumstances when you actually need it. At all other times, nobody cares anyways. So economically I don't see the difference to the depositor between holding CB deposits and holding BMO deposits.
Posted by: K | April 03, 2012 at 04:36 PM
"Macroeconomists generally believe that the Bank of Canada would then control monetary policy both for Canada and for the US. "
Nick,
Ben Bernanke says the same about China.
I thought the PBoC sets its own interest rate.
Posted by: Ramanan | April 03, 2012 at 04:38 PM
1. and 2.
Suppose Bank of Canada keeps CPI expectations in the range of 1.8%..2.2%. In this case both Bank of Montreal and Fed could manipulate expectations inside that range, thus affecting the monetary policy.
Suppose Bank of Montreal is deleveraging and is pushing CPI expectations to the bottom of the range, while Fed is doing the opposite to fight the recession in the US. Infinite force meets the immovable object...
The same logic applies to the fiscal policy. Taxes should be cut until Bernanke promises to start raising interest rates three months earlier.
Posted by: 123 | April 03, 2012 at 04:39 PM
Ramanan: China has capital controls though? OK, since there aren't any capital controls inside Canada, I would have to assume there are no capital controls between Canada and the US (roughly right, I think).
123. OK. I expect we could say that BMO has a little bit of control within Canada too. If it suddenly went wild to expand or contract, it might take a bit of time for the BoC to react and offset it.
Posted by: Nick Rowe | April 03, 2012 at 05:01 PM
Nick
Btw, did you get my email?
Posted by: Ritwik | April 03, 2012 at 05:19 PM
Ritwik: Yes. I just checked. Will respond later, when my brain is less fried!
Posted by: Nick Rowe | April 03, 2012 at 05:46 PM
Geez Nick! This is like 5 posts since March 30th. Don't you have a day job?
Anyway, the answer to your question (from my own, ahem, unique perspective) is that convertibility is irrelevant to the value of either kind of dollar. If the Fed has assets worth 100 oz. of silver backing $100, then each Fed dollar will be worth 1 oz of silver whether or not the Fed maintains silver convertibility. If the BoC has assets worth 100 oz of gold backing 100 Canadian dollars, then each Canadian dollar will be worth 1 ounce of gold, regardless of gold or US dollar convertibility.
Posted by: Mike Sproul | April 03, 2012 at 05:59 PM
maybe its the "assymmetric redeemability" nomenclature that scares people away.
Seems to me, you have to look no further than Europe right now to find a near exact example, since all 27 countries (which all have their own CB and national banking systems) are all effectively tied to the Euro (and as I understand it, each CB even issues its own "Euros" with serial numbers ending in a letter specific to that country) with the Euro-wide "Target2" payment system. Does anyone really doubt the ECB, not the Greek central bank, controls monetary policy in greece?
Posted by: dwb | April 03, 2012 at 06:04 PM
Mike: Yep! I need to slow down, and take a break, to restore my sanity (such as it is).
I disagree with you, of course, as always! (Yours is almost unique, AFAIK). What if the cross-rate between the US dollar and the loonie contradicts the cross rate between gold and silver, so there's arbitrage profits? Are we back into the bimetallic controversies?
dwb. The Euro analogy is interesting, but I'm not sure it works in this case. What would be the equivalent to Target2 between the BoC and BMO? And between the Fed and the BoC?
Posted by: Nick Rowe | April 03, 2012 at 06:27 PM
Nick: Yes; it would be a really bad idea for either bank to set a conversion rate between the US$ and the loonie, since any difference between that and the market rate puts the pegging bank on the wrong end of an arbitrage process.
Posted by: Mike Sproul | April 03, 2012 at 06:41 PM
I also believe asymmetric redeem ability is a very important factor and is one crucial aspect of understanding the special role of the central bank's money in the hierarchy of credit and a sovereign monetary system. But I would be cautious in saying this fact by itself gives the central bank "control" over over the monetary system. The fact that one kind of monetary instrument or IOU is officially and legally redeemable in some base money doesn't mean that much in itself. The IOU's, if freely issued, can take on a monetary life of their own and become an nearly independent medium of exchange. It might come about that they are rarely redeemed for base money, and so the quantity of base money in the possession of the IOU issuer comes to have dwindling importance to the issuer and the holder of the IOUs. As a result, it might come about that decisions by the central bank that issues the base money also have correspondingly less significance and impact.
Other factors must pay a role if the IOUs are to stay closely ties to the base into which they are redeemable. in the US, for example, it's not just that commercial bank money is legally redeemable in US currency that counts, but that the payments system is organized in a hierarchical manner so that routine interbank payments actually are actually made with Fed reserve balances. In other words, the Fed's money in the form of reserve balances is the universal medium of exchange for interbank payments. That gives the Fed a much more important role to play. It can't always do much to boost the expansion of credit by providing more reserves - especially at the zero bound; but it could certainly substantially reduce the expansion of credit if it for some reason decided to restrict severely the provision of reserve balances.
To other factors play a role in giving the government's money a powerful, monopoly role in the economy: the decision by a government to impose tax liabilities that can only be discharged using the monetary base; and host's of jurisprudential arrangements establishing the base money as a final means of payment. Note that when a citizen pays taxes in the US, that payment is ultimately settled and cleared between the citizen's bank and an account with the treasury. The citizen's account at the Fed is debited by some quantity; and the same quantity is credited to a treasury account.
If the government accepted the commercial bank's IOU as such in payment of a tax bill, in the same way private sector individuals accept the IOU of their bank from another customer of the same bank, without any payment being made out of the bank's reserve account, then the significance of the reserve account - and thus the government's base money - would be lost. We would be on our way to a free banking system in which commercial banks each issued their own private "fiat" money. These different monies might be mutually accepted among different banks, but it would be similar to the way in which one sovereign government accepts the fiat money of another sovereign government. It would be a case of pure convenience for the purposes of commerce, without any formal guarantee of a common basis of redemption.
Posted by: Dan Kervick | April 03, 2012 at 07:05 PM
Sorry about all the typos. My do needs a walk and I was typing fast!
Posted by: Dan Kervick | April 03, 2012 at 07:07 PM
IMO:
The Bank of Montreal dollars issues liabilities which are denominated into Bank of Canada dollars, the Bank of Canada makes sure via institutional arrangements that the bank of montreal is always able to 'convert' the BoM's liabilities into dollars. I really don't understand the issue that you raise here? Institutional arrangements ensure that BoM liabilities are readily convertible into BoC liabilities.
If the USD was pegged to CAD, and if the CAD made no commitments, then this is a fundamentally different institutional arrangement to the banking system of Canada. US monetary policy would be required to defend the peg, and ensure that it had sufficient CAD to defend its peg, as well as using monetary policy to control CAD in flows.
Posted by: mdm | April 03, 2012 at 07:28 PM
Dan: Hope your dog wasn't crossing its legs!
"The IOU's, if freely issued, can take on a monetary life of their own and become an nearly independent medium of exchange. It might come about that they are rarely redeemed for base money, and so the quantity of base money in the possession of the IOU issuer comes to have dwindling importance to the issuer and the holder of the IOUs. As a result, it might come about that decisions by the central bank that issues the base money also have correspondingly less significance and impact."
But the fact that they *could be* redeemed means they will always have the same value, and (ignoring risk) interest rates measured in the two monies would also have to be the same.
mdm: "US monetary policy would be required to defend the peg, and ensure that it had sufficient CAD to defend its peg, as well as using monetary policy to control CAD in flows."
Let me rewrite that, with some changes:
'BoM monetary policy would be required to defend the peg, and ensure that it had sufficient BoC dollars to defend its peg, as well as using monetary policy to control BoC in flows.'
I think it still works, doesn't it? That's the issue I'm raising.
Posted by: Nick Rowe | April 03, 2012 at 07:57 PM
The Euro analogy is interesting, but I'm not sure it works in this case. What would be the equivalent to Target2 between the BoC and BMO? And between the Fed and the BoC?
maybe i am not getting what you are driving at. The way i see it, the fundamental issue is which payments/settlements/liabilities have to be settled in "central bank money," and which do not. Cenntral bank money is special because its redeemable in currency, the central bank has a monopoly, and there is an infinite supply. The flows around target2 for example are all central bank money so if an Irish Farmer buys a German tractor, the Irish bank settles with the German bank in "central bank money" to effect the payment. the BoC settles accounts with BoM in "central bank money" but not with the Fed (unless there is a truly blessed monetary union!). The liabilities of the BoM (my deposits) are just liabilities of them, so if they go under, poof (unless they are TBTF)! The Fed has no claims against the BoC, even if they've pegged the dollar.
Liabilities of different institutions are generally not fungible, even if they are in the same unit/currency; But "central bank liabilities" or money is special.
Posted by: dwb | April 03, 2012 at 09:13 PM
But the fact that they *could be* redeemed means they will always have the same value, and (ignoring risk) interest rates measured in the two monies would also have to be the same.
I'm not sure, Nick. The fact that bank IOUs are redeemable does mean that the value of the central bank's money would set a floor on the value of the bank IOUs. The bank money could never fall below the redemption rate since people would rush to redeem the banks' notes if that happened. But would it necessarily have to work the other way? If the value of the bank IOUs floated far above the value of the redemption quantity of the central bank base, then no would would ever in fact redeem the IOUs, even though they were redeemable.
That's not inconceivable. We could imagine a system in which we all use bank notes that are inscribed with a solemn and legally enforceable promise permitting you to redeem the notes in federal garbanzo beans - a special hybrid strain of garbanzo beans that are grown and stored at a few government farms. Garbanzo beans are also the standard unit of account, but imagine the market value of the notes is now such that when you buy an iPad, the going price is about a ten garbanzo note. If you want, you can always redeem a ten garbanzo note for ten actual government garbanzo beans. But of course, nobody would ever actually do this because for most people ten garbanzo beans are worth far less than an iPad.
All we could say for sure is that the ten garbanzo bean note would never drop below the value of ten actual garbanzo beans, because then people would redeem them.
It seems to me that what gives a central bank and government important monopoly control over the value of the nation's broader currency in a fiat monetary system is not just the offer of redeemability into the base, but the fact that in various ways people are compelled to transact business with the government's money and accept that money, and also that the government does various things to establish a value for that base money when it offers it in exchange for goods and services and accepts it in exchange for goods and services. On the chartalist perspective, one of these actions is the requirement that you obtain the money to discharge a tax obligation. If you were required to pay a hundred garbanzo beans each year in taxes, and the penalty for failure to pay were one day in jail for each unpaid bean, then the market value of garbanzo beans would quickly rise above their previous commodity value as edible beans, and come to be equal to the value of one day of liberty per bean.
And if the government runs a central banking system in which banks are required to settle their payment obligations among each other in actual garbanzo beans, at par in the standard unit of account, that that would bring the value of the ten garbanzo bean note in line with the value of ten actual garbanzo beans. If Bank A's depositor offered Bank B a check on his account for Ten Garbanzo Beans, in the unit of account corresponding to the value of one of the Ten Garbanzo Bean notes issued by Bank A, and Bank B knew that what that meant is that the payment would be settled and cleared by the delivery of ten actual garbanzo beans, then Bank B would never accept the check unless there weren't much difference between the market value of the Ten Garbanzo Bean note (the bank money) and ten actual garbanzo beans (the government's base money).
So it seems to me you need to add something to asymmetric redeemability for the value of the government's base money to act as both a floor and a ceiling.
Posted by: Dan Kervick | April 03, 2012 at 10:09 PM
My intuition tells me it wouldn't matter in the least. So what if Canada is a smaller monetary entity than the US?
And based on Canada's relative eonomic performance it might be an improvement if the US pegged the US dollar to the loonie.
Posted by: Mark A. Sadowski | April 03, 2012 at 10:20 PM
Just wanted to say that I'm finding this discussion to be truly elucidating and to thank Mr. Rowe for being such a good sport. It can be maddening getting in a discussion like this and I (and I'm sure most of your readers) really appreciate the effort and patience you're showing. I doubt anyone is going to change their position, but that doesn't mean minds aren't being influenced.
Posted by: Simon | April 04, 2012 at 03:15 AM
Sorry, I meant to comment on the prior post. I had both of them open in different tabs, oops!
Posted by: Simon | April 04, 2012 at 03:35 AM
Nick,
Your analogy does seem right mechanically. If the BoC raises (lowers) rates, the Fed will be obliged also to raise (lower) rates, or else it will run out of CAD to be redeemed (be flooded with money.) Substitute BMO for the Fed, it works the same.
So I have been asking myself why I resist your figure. I think that K's first observation gets to the heart of it: institutional arrangements are such that there is almost no risk that a BMO dollar will not be redeemed for a BoC dollar. The BoC does indeed exercise a powerful influence over BMO's "monetary policy", but it does so deliberately and directly, via means that do not seem directly connected to concerns over convertibility. In your thought experiment, the BoC would control the Fed's policy accidentally via means that are directly connected to concerns over convertibility. Despite the coincidence in forms, in interpretation the two cases are almost opposite.
Posted by: Phil Koop | April 04, 2012 at 07:07 AM
Analogy:
Suppose monetary systems are computer operating systems.
Canada is Mac's OSX.
The US is Microsoft's Windows.
BOC is the root system, while the BOM is a programme that uses the BOC's language as its logical structure.
BofA is a programme that runs on the Fed's Windows system. But the Fed has mandated that its own underlying dynamics mimic those of the BOC. The language is still Windows, but the dynamics mimic the Canadian system and the BofA has no choice but to follow suit.
Even if the Fed decides to use the BOC's moentary policy as an exogenous reaction funtion, the clearing system is still maintained by the Fed, not the BOC and US$ remain US$, whereas CAD remain CAD.
This is manifest in the distinct possibility of the Fed to end the peg, as opposed to the BOM, that can not (realistically) make such a decision. Monetary hierarchy is a legal hierarchy and legal power wrt monetary policy usually in the hands of the national central bank and not a subset (commercial bank) or a foreign entity.
Make sense?
Posted by: Oliver | April 04, 2012 at 07:22 AM
I'll try a scenario to see if it helps.
Let's say for whatever reason the BoC had an extremely tight stance and so the loonie is scarce. In contrast at the time the peg is established the Fed has a much looser stance. So BoM struggles to attract funds at a sufficiently low rate to grow. Citibank has a very low cost of funds. An arbitrage exists which drives Citi to lend to BoM in the interbank market and then redeem with the Fed. Then the fed must settle at par. The fed has a problem controlling inflation.
So yes the BoC controls monetary policy of the USA in this example, even though the US is much bigger.
Doesn't Germany have this problem now too?
Posted by: D | April 04, 2012 at 07:24 AM
Jus read Ritwik's link from above. I find my example fits quite well with it.
Posted by: Oliver | April 04, 2012 at 07:27 AM
Nick and K:
But the Bank of Canada doesn't guarantee bank deposits. The CDIC does. In the event of serious crises, making sure that the CDIC has the cash to reimburse depositors is the Fed's government's problem, not the BoC's. Note that during the crisis in 2008-2009, the FDIC (US counterpart of CDIC) basically burned through its entire reserve fund in a few quarters and started to be very vague on the state of its reserves fund and expected claims on it (all the while claiming that depositors had no reason to fear.)
There's another important difference. Deposit insurance insures *small* deposits; not enough enough to meet the weekly payroll for a good-sized business, much less a financial firm. When things look dicey, large accounts still flee. (Reuters was reporting late last year that several major European corporations, fearful of a banking crisis, were taking cash out of the commercial banking system altogether and instead depositing directly with the ECB.)
But as Nick says, deposit insurance is rarely tested.
Posted by: Simon Van Norden | April 04, 2012 at 10:01 AM
Most people thinking about fixed exchange rates (if that's not the same as Asymmetric Redeemability, please correct me, Nick!) got an important smack up-side the head after the Argentine crisis a decade ago. Argentina ran a currency board to protect its peg to the US dollar. (This meant that for every $1 worth of Argentine currency they issued, they had $1 USD in reserves.) Until their crisis, most economists believed that this arrangement was speculation-proof and provided an iron-clad link between the two currencies. However, it lasted only a few years before failing disastrously.
One lesson that many drew from that experience is that provisions for behaviour in extreme and unlikely scenarios are surprisingly important in understanding how monetary systems will work and how likely they can last.
So what if the Fed were to peg to the loonie? I'd argue that two factors would matter a lot to monetary conditions in the US.
1) BoC policy
2) The extent to which people think that the Fed will keep the peg. That depends both on (a) whether they have the means to do it, and (b) whether they have the will to do it.
Posted by: Simon Van Norden | April 04, 2012 at 10:20 AM
Oliver: I'm not sure your analogy works here, but I think it's going to work great in a slightly different context. So, going slightly off-topic (hey, it's my post!) can anyone help me with (I'm not sure if my questions make sense):
1. What's the name of a common really really low level machine language?
2. What's the name of an old-fashioned defunct mid-level programming language? Cobol?
3. What's the name of a common modern mid-level programming language? C+ ?
4. What's the name of a common modern upper-level language for users? Windows?
The interest rate story is about how C+ runs windows. The people who tell the interest rate story think the textbook multiplier story is told in Cobol, but it's really about how machine language works to run windows, skipping C+.
Posted by: Nick Rowe | April 04, 2012 at 11:02 AM
Simon: "Most people thinking about fixed exchange rates (if that's not the same as Asymmetric Redeemability, please correct me, Nick!).."
It's almost always the same. I can only think of one exception, AFAIK. France and Germany immediately before the EURO was introduced. That was *symmetric* redeemability, because both countries undertook a joint responsibility to defend the DM/Franc fixed exchange rate and (presumably) consulted with each other and came to a joint decision on monetary policy.
"2) The extent to which people think that the Fed will keep the peg. That depends both on (a) whether they have the means to do it, and (b) whether they have the will to do it."
and Phil:"I think that K's first observation gets to the heart of it: institutional arrangements are such that there is almost no risk that a BMO dollar will not be redeemed for a BoC dollar."
Yep. I think you're right, and that's important. There's always the chance the Fed will break the peg, and yet the USD would keep going as a medium of exchange. If BMO broke the peg, I'm not at all sure that BMO dollars would keep being used as a medium of exchange, trading at a discount. I think some Scottish banknotes did continue to trade at a discount to gold in the olden days. I wonder if there's a difference there between paper currency and demand deposits??
Posted by: Nick Rowe | April 04, 2012 at 11:14 AM
The *assembly* language for your PC/Mac whatever is called x86. It's not literally machine code, but it's the human equivalent thereof. Assembly code is CPU architecture specific.
C is a fairly simple "mid-level" language which can be compiled (to assembly code) and assembled to machine code for a specific machine architecture.
C++ is an extension of C that has lots of higher level features that don't map in any simple way to underlying assembly code.
Windows is not a language. It's an operating system. You don't write a program in Windows. You write it in C++, .NET or whatever.
To complete you analogy:
The operating system is the economy.
The path of the policy rate is the machine language.
The money supply is Cobol: an irrelevant vestigial appendage.
Is that what you meant?
Posted by: K | April 04, 2012 at 11:25 AM
Dan: "If the value of the bank IOUs floated far above the value of the redemption quantity of the central bank base, then no would would ever in fact redeem the IOUs, even though they were redeemable."
Hmmm. Good question. Why does that never happen in practice? I think the answer is that if it started to happen, so that people preferred to own BMO dollars rather than BoC dollars, people would immediately rush to the BMO to deposit their currency, and BMO would accept that currency, and open an account. There's no downside to BMO, since it can always sit on that currency if it wants, and charge a fee to cover the costs of maintaining the demand deposit?
"That's not inconceivable. We could imagine a system in which we all use bank notes that are inscribed with a solemn and legally enforceable promise permitting you to redeem the notes in federal garbanzo beans..."
Yep, but in that case BMO dollars really would be a totally different currency to BoC dollars, with a floating exchange rate.
Posted by: Nick Rowe | April 04, 2012 at 11:37 AM
Phil Koop said,
"If the BoC raises (lowers) rates, the Fed will be obliged also to raise (lower) rates, or else it will run out of CAD to be redeemed (be flooded with money.) Substitute BMO for the Fed, it works the same."
It's an interesting observation. I have been always wandering of the exact mechanism that drives commercial banks to follow central banks in their interest rate policy.
Let me try to devise an arbitrage here based on Phil's observation.
BoC decreases rates and BoM doesn't. Then other banks borrow money from BoC and deposit them in BoM. BoM goes bankrupt, so it has to decrease its rates.
BoC increases rates and BoM doesn't. Then other banks borrow money from BoM instead of BoC when they need a loan. BoM goes bankrupt, so it has to increase its rates.
It sounds like a solid arbitrage mechanism to me. Please criticize.
Posted by: Alex | April 04, 2012 at 12:41 PM
Dan Kervick said,
"If the value of the bank IOUs floated far above the value of the redemption quantity of the central bank base, then no would would ever in fact redeem the IOUs, even though they were redeemable."
This is what happens in the EZ. People withdraw money from peripheral banks and deposit them to German banks.
Posted by: Alex | April 04, 2012 at 12:56 PM
Alex: your arbitrage mechanism sounds right to me. It's just interest rate parity, under fixed exchange rates, the Mundell-Fleming BP curve in a different context.
Posted by: Nick Rowe | April 04, 2012 at 01:05 PM
Nick,
Thanks. Actually, I was thinking that CBs (forcefully) persuade commercial banks to adjust interest rates since I have not seen the exact mechanism before.
Posted by: Alex | April 04, 2012 at 01:22 PM
Alex,
Yes, by arbitrage there can only be one risk-free rate. The only question is, who has the power to determine it. The answer is the CB because they have infinite capacity both to lend and to borrow. Therefore they are Chuck Norris: they only have to say what the rate will be. Since, at the end of the day, that's what the rate will be, if any overnight transactions are done at some other rate, one of the parties to that trade will have gotten screwed. Therefore all trades get done at the policy rate. So long as all interbank balances are cleared by the end of trading, no reserve balances (or discount window borrowing) are required.
Posted by: K | April 04, 2012 at 01:25 PM
"1. If that were to happen, in what important ways would the US Fed be different from what the Bank of Montreal is now?"
I'd say that the differences between the Fed and BMO given your scenario would be that:
1. the Fed would continue to act a formal lender of last resort to the US banking system while simultaneously pegging to the c-dollar, whereas BMO would only be pegging to the c-dollar, it wouldn't have a formal LOLR role to the Canadian banking system. (Even this is not entirely correct since BMO is contracted to provide back up lines of credit, liquidity backups etc. to some of its clients).
2. in pegging to the c-dollar, BMO gets to be a member of the Canadian Payments Association and a participant in the LVTS, it has a Bank of Canada account, and would be able to borrow from the BoC and participate in open market operations knowing that at x% the BoC is willing to create an infinite amount of c-dollar reserves on good collateral. Even though it is also pegging to the c-dollar, the Fed wouldn't have these benefits. Maybe a limited swap line.
Posted by: JP Koning | April 04, 2012 at 01:30 PM
Nick,
This loonie peg story illustrates, that it is the capital, and not irredeemability that is the source of the monetary power. Suppose it is believed that Bank of Canada keeps CPI expectations in the range of 1.8%..2.2%. And then the Fed creates Canadian CPI futures a-la-Sumner, and pegs them at 3%. The central bank with the deepest pockets wins. Irredeemability franchise is just one of the inputs into the calculation of the economic capital of central bank.
"I expect we could say that BMO has a little bit of control within Canada too."
Yes. Because BMO has a tiny capital cushion, the control is very limited.
Experiment 2. Suppose Canadian treasury creates Canadian CPI futures market, and pegs them at 3%. Suddenly ECB treaty prohibition on monetary financing of budget deficits is looking much better.
Posted by: 123 | April 04, 2012 at 01:34 PM
K,
I agree in principle but I disagree in details. It is indeed the reserve balances that are playing a crucial role here.
Commercial bank goes bankrupt when he runs out of reserve balances -- had a nice discussion on them lately :-) -- with the CB. It is like running out of gold in the ancient times.
Posted by: Alex | April 04, 2012 at 01:38 PM
Alex,
Reserve balances have nothing to do with bankruptcy. A bank goes broke when they are not able to settle interbank balances. If they have *eligible* collateral, they can just borrow reserves at the discount window. Canadian banks can easily have negative reserve balances. Even the Fed has a long term project to eliminate net reserves (so often some banks would be negative). So long as banks have adequate capital and have liquid good collateral, nobody cares. What determines the event of bankruptcy is the lack of eligible collateral.
Posted by: K | April 04, 2012 at 01:51 PM
K,
Yep, I am actually talking about the same. Banks use reserve balances for interbank transactions. When they can't replenish them they go bankrupt.
Posted by: Alex | April 04, 2012 at 01:59 PM
Alex,
OK. So what were you disagreeing with? What I said was that overnight "no reserve balances (or discount window borrowing) are required." I.e. Since there are extra costs with a non-zero reserve balance (either positive or negative), under normal conditions all banks have zero reserves. Then you said that that would make them bankrupt. Are you not saying that any more?
Posted by: K | April 04, 2012 at 02:25 PM
Me,
"Commercial bank goes bankrupt when he runs out of reserve balances with the CB. It is like running out of gold in the ancient times."
Actually I want to elaborate on this to show my respect to Nick's site.
In the ancient time: intrabanks transactions were settled in gold. Client got IOU from Bank A, came to Bank B and deposited it or cashed it out. Bank B sent IOU to Bank A and received payment in gold. When bank run out of gold it went bankrupt.
Now: intrabank transactions are settled out in "reserve balances." Client gets IOU from Bank A, comes to Bank B and deposits it or cashes it out. Bank B sends IOU to Bank A and receives payment in "reserve balances." When bank runs out of "reserve balances" it goes bankrupt.
Posted by: Alex | April 04, 2012 at 02:28 PM
K,
I didn't like this phrase "no reserve balances are required."
This is why I said I disagree in details. Even in Canada they are implicitly required to settle intrabank transactions.
Posted by: Alex | April 04, 2012 at 02:32 PM
Alex,
You are really messing with the language. A balance is a position on a balance sheet. Your reserve position with the central bank can be in positive or negative *balance*. You don't "run out of reserve balances". Having a zero reserve balance is a natural and desirable balance sheet position (in Canada). You can however run out of the good collateral required in order to borrow in the interbank market the *reserves* required to settle your interbank payments. Then you can try to use your crappy collateral to borrow said reserves from the CB (thereby giving you a negative reserve balance) to meet your obligations in the payment system.
If nobody is willing to lend to you, I guess you could say that you ran out of *reserves*, though it would be a weird thing to say since your reserve *balance* probably went negative a long time ago. You didn't run out, you went way negative. What makes sense to say is that you *ran out* of payment system eligible collateral and as a result you were no longer able to borrow any more reserves.
"I didn't like this phrase "no reserve balances are required.""
This just means that Canadian banks neither borrow from, nor lend to the CB on a normal basis. US banks own the mandatory reserves which they keep on deposit at (lend to) the Fed.
"Even in Canada they are implicitly required to settle intrabank transactions."
Actually, they are *explicitly* required to settle interbank transactions. I.e. they must absolutely not be short at the end of the day. If a bank is in danger of failing to pay they have to run to the BoC for a loan before market close, something funding desks really try to avoid.
Posted by: K | April 04, 2012 at 03:44 PM
Hmmm. Good question. Why does that never happen in practice.
I think the chartalist mechanisms I mentioned have something to do with it, Nick. Governments make laws and have real powers. They can do things that make it difficult for people not to transact business in the state's money, and create strong disincentives for using alternative currencies. If they set the one-way exchange rate between currency X and the state currency by fiat, and create strong demand for the state currency by establishing legal obligations that require you to acquire the state's currency in certain specific quantities to discharge the obligations, then they are going to have a lot of control over the other-way exchange rate.
Posted by: Dan Kervick | April 04, 2012 at 03:57 PM
K,
I don't want to argue about semantics. I just tried to say that even if a bank doesn't have reserve balances at the moment it is implicitly assumed that a solvent bank will get them (from any source) when there is time to settle.
Perhaps you had not explained your idea clearly previously, so I said I didn't like it. It's no point to argue anymore as I see it.
Posted by: Alex | April 04, 2012 at 04:15 PM
Nick; Thanks for the clarification.
BTW, the EMS (which included the DM/FF example that you mentioned) was symmetric de jure but asymmetric de facto; the Bundesbank never loosened its monetary to avoid putting strain on the currency of member states. The central banks "consulted" one another in the same sense that the Bank of Canada and the Dept. of Finance "consult regularly"; many governors have explained that they interpret this to mean simply that the Minister of Finance should never be taken by surprise by the Bank's actions.
Posted by: Simon van Norden | April 04, 2012 at 08:32 PM