Commercial banks are typically beta banks, and central banks are typically alpha banks. Beta banks promise to convert their money into the money of alpha banks at a fixed exchange rate. Alpha banks make no such promise the other way. It's asymmetric redeemability. This means there cannot be an excess supply of beta money in terms of alpha money. (Nor can there be an excess demand for alpha money in terms of beta money.) Because people would convert their beta money into alpha money if there were. But there can be an excess supply of beta money in terms of goods, just as there can be an excess supply of alpha money in terms of goods. If beta money is in excess supply in terms of goods, so is alpha money, and vice versa. If commercial and central bank monies are perfect or imperfect substitutes, an increased supply of commercial bank money will create an excess supply of both monies against goods. The Law of Reflux will not prevent this.
This is in response to David Glasner's good post. David has forced me to be much clearer about what it means to say there is an excess supply of commercial bank money.
Money, the medium of exchange, is not like other goods, because if there are n goods plus one money, there are n markets in which money is traded, and n different excess supplies of money. Money might be in excess supply in the apple market, and in excess demand in the banana market.
If there are two monies, and n other goods, there are n markets in which money is traded against goods, plus one market in which the two monies are traded for each other. If beta money is convertible into alpha money, there can never be an excess supply of beta money in the one market where beta money is traded for alpha money. But there can be an excess supply of both beta and alpha money in each or all of the other n markets.
Start in equilibrium, where the existing stocks of both alpha and beta money are willingly held. Hold constant the stock of alpha money. Now suppose the issuers of beta money create more beta money. Could this cause an excess supply of money and an increase in the price level?
If alpha and beta money were perfect substitutes for each other, people would be indifferent about the proportions of alpha to beta monies they held. The desired share or ratio of alpha/beta money would be indeterminate, but the desired total of alpha+beta money would still be well-defined. If beta banks issued more beta money, holding constant the stock of alpha money, the total stock of money would be higher than desired, and there would be an excess supply of both monies against all other goods. But no individual would choose to go to the beta bank to convert his beta money into alpha money, because, by assumption, he doesn't care about the share of alpha/beta money he holds. The Law of Reflux will not work to eliminate the excess supply of alpha+beta money against all other goods.
Now suppose that alpha and beta money are close but imperfect substitutes. If beta banks want to prevent the Law of Reflux from reducing the stock of beta money, they would need to make beta money slightly more attractive to hold relative to alpha money. Suppose they do that, by paying slightly higher interest on beta money. This ensures that the desired share of alpha/beta money equals the actual share. No individual wants to reduce his share of beta/alpha money. But there will be an excess supply of both alpha and beta monies against all other goods. If apples and pears are substitutes, an increased supply of pears reduces the demand for apples.
Now suppose that alpha and beta money are neither substitutes nor complements. Only in this case would an increased stock of beta money, plus a large enough increase in the rate of interest paid on beta money to ensure that nobody wants to convert beta money into alpha money, mean that there is no excess supply of alpha and beta money in terms of goods. If Canadians use only alpha money, and Americans use only beta money, and the Fed promises to convert US beta dollars into Canadian alpha dollars at a fixed exchange rate, the Fed could not create inflation by increasing the supply of beta money and making it more attractive to hold.
Purely for completeness, if alpha and beta monies were complements, if beta banks increased the stock of beta money, and increased the rate of interest paid on beta money to ensure nobody wanted to convert beta money into alpha money, this would create an excess demand for both alpha and beta money against all other goods.
Commercial banks are beta banks. They fix the exchange rates of their monies against central bank alpha money. It seems reasonable to assume that commercial bank money is an imperfect substitute for central bank money. Chequing accounts are an imperfect substitute for currency. If so, an increase in the supply of commercial bank money, with commercial banks taking actions to ensure that people want to hold a larger share of commercial bank money relative to central bank money, will create an excess supply of both monies against all other goods.
The Law of Reflux would work very well for any good except money. If producers of refrigerators stand ready to convert refrigerators into money or vice versa at a fixed exchange rate, there could never be an excess supply or demand of refrigerators in terms of money. Nor could there ever be an excess demand or supply of money in terms of refrigerators. But refrigerators are traded in one market, for money. Money is traded in all markets, for all other goods. There could never be an excess supply or demand of money in terms of refrigerators; but there could still be an excess supply or demand of money in terms of everything else.
Nick,
As preamble, I’d still like to be clearer on your asymmetric redeemability argument.
a) The central bank sells BMO currency in exchange for a BMO reserve balance debit, and BMO sells its customer currency in exchange for BMO deposit debit. Customer walks away with currency and no BMO deposit. BMO walks away with lower reserves and deposits.
b) The central bank sells BMO a reserve balance credit in exchange for currency, and BMO sells its customer a deposit balance credit in exchange for currency. Customer walks away with a BMO deposit and no currency. BMO walks away with higher reserves and deposits.
Would you agree that’s symmetric?
How does your asymmetry argument progress from that point?
Posted by: JKH | March 28, 2014 at 10:04 AM
JKH:
a. there are two types of central bank money: currency and reserves. The central bank promises to convert either into the other at a fixed exchange rate. This is no different from converting $20 notes into two $10 notes. We can ignore this.
b. Who fixes the exchange rate between BMO money and BoC money? I say it's BMO. I can't walk into the BoC, give them my BMO cheque for $100, and demand $100 in BoC notes. I can walk into BMO and do that.
Posted by: Nick Rowe | March 28, 2014 at 10:17 AM
"I can't walk into the BoC, give them my BMO cheque for $100, and demand $100 in BoC notes."
That's an odd example in terms of "my BMO cheque"
Suppose we both bank at BMO.
I write a check to you.
You cash it at BMO in exchange for currency.
The fact that you can't cash it the Bank of Canada is an institutional clearing constraint in my view - not an indication of a redeemability characteristic.
The Bank of Canada clears BMO cheques all the time in exchange for reserves/currency. That's the essential redemption action - not whether you go to BMO or Royal or the central bank to get the currency.
BMO is just the clearing agent for currency.
The check gets redeemed for currency regardless of the clearing mechanics and who is acting as clearing agent for whom.
I'm puzzled by what this example demonstrates.
And I'm puzzled why you chose the direction of redemption into currency rather than vice versa.
Posted by: JKH | March 28, 2014 at 10:37 AM
Nick,
What I’m not getting more generally is that BMO can destroy or create its beta money in exchange for the creation or destruction of alpha money - in its capacity as distribution agent for alpha money.
I see no constraint on either alpha efflux (beta reflux) or alpha reflux (beta efflux) – except for the supply of beta reflux (i.e. BMO deposits available to be redeemed) and the demand for beta efflux (demand for new BMO deposits). I don’t see that constraint being one of asymmetric redeemability.
Posted by: JKH | March 28, 2014 at 10:49 AM
"not getting" meaning that's the way I see it
Posted by: JKH | March 28, 2014 at 10:50 AM
I probably bungled your explanation of asymmetric redeemability as the defining feature of central banks when I attempted to explain it to Steve Williamson, but he did at least respond.
Posted by: Wonks Anonymous | March 28, 2014 at 10:53 AM
JKH: suppose BoC promised to convert BMO money into BoC money at par, regardless. So the BoC is the one that pegs the exchange rate. If I were in charge of BMO, I would then print BMO money like crazy, to make lots of profits, and if my customers wanted BoC money I would just get more from the BoC and give it to them. Or, maybe I would be public-spirited and tell all Canadians that we are now targeting NGDP, and not 2% inflation.
Posted by: Nick Rowe | March 28, 2014 at 10:54 AM
Wonks: I saw that. You didn't bungle it. His response added some minor qualifications, which sounded OK to me. In particular, his bit about it depending on what the alpha bank is targeting is correct. But the alpha bank is the one that gets to choose the target.
Posted by: Nick Rowe | March 28, 2014 at 11:03 AM
Nick,
"suppose BoC promised to convert BMO money into BoC money at par, regardless"
It does that.
So long is BMO is up and running as a bank.
If BMO isn't up and running, its because of bad risk management, and regulatory clamp down.
But I fail to see how that is indicative of asymmetric redeemability - that's just shutting down a bank.
Posted by: JKH | March 28, 2014 at 11:03 AM
Nick,
Did you see my comments at PragCap - distinguishing between the denominator (alpha insurance) and the numerator (credit risk insurance)?
Asymmetric redeemability is about alpha insurance - not credit risk insurance.
http://pragcap.com/lets-talk-about-seigniorage
Posted by: JKH | March 28, 2014 at 11:06 AM
i.e. where credit risk insurance in this case means deposit insurance - same diff
Posted by: JKH | March 28, 2014 at 11:08 AM
Nick,
"Commercial banks are beta banks. They fix the exchange rates of their monies against central bank alpha money."
Commercial banks are actually debtors or borrowers, rather than 'currency issuers' in a sort of quasi-foreign exchange market. A bank deposit is bank debt denominated in the currency issued by the central bank.
Many US banks state this explicitly in their account terms and conditions. The usual phrase used is: "our deposit relationship with you is that of debtor and creditor".
Posted by: Philippe | March 28, 2014 at 11:23 AM
The reason a bank deposit "exchanges at par" with central bank money is because legally a bank deposit represents a specific amount of central bank money borrowed by the bank from the depositor.
Posted by: Philippe | March 28, 2014 at 11:39 AM
Nick,
My theory now is that I agree with what you are saying, but think that what you are saying is captured best in a framework other than “asymmetric redeemability”.
The only asymmetry I see is that over a long period of time, both alpha and beta money will tend to grow rather than contact – due to NGDP growth.
So there is a net asymmetry in term of more cumulative redemption of beta money for alpha over time.
That’s just a function of growth.
But there is no asymmetry in terms of the option to redeem in either direction at any time.
Unless a bank is shut down – then all bets are off.
The regulatory framework should aim to ensure that banking is efficient and effective in all of its aspects.
And that it is compatible with monetary policy flexibility.
So – symmetric redemption options at all times; long term asymmetry toward net growth with cumulative net redemption of beta for alpha. That cumulative growth amount will be steered by monetary policy.
Posted by: JKH | March 28, 2014 at 11:39 AM
JKH and Philippe: forget all the institutional details, because it only confuses things.
The BoC prints bits of paper with "BoC" written on them, and nothing else. All bits of paper are identical.
The BMO prints bits of paper with "BMO" written on them, and nothing else. All bits of paper are identical.
What determines the exchange rate between the two types of paper. It could be anything. There is nothing to say it should be one-for-one.
If BMO and BoC try to fix two different exchange rates, who wins?
Which one will adjust its exchange rate to the other's?
"BMO" and "BoC" are just two arbitrary names, of two arbitrarily chosen banks. We don't know which of the two banks can set the monetary policy target, until we have answered the question of who sets the exchange rate.
The one which does not choose the exchange rate gets to choose whether to target inflation or NGDP. Standard open-economy macro. You can't fix the exchange rate and have an independent monetary policy at the same time. If the other guy fixes the exchange rate, you get to target what you want for both of you.
Posted by: Nick Rowe | March 28, 2014 at 11:57 AM
“Asymmetric redeemability" is just another way of stating the fact that that banks borrow money from depositors.
You can exchange your bank deposit for central bank money, because that is what the bank has borrowed from you and promised to pay you.
When you deposit central bank money at a bank, or exchange central bank money for a bank deposit, the deposit-issuing bank goes into debt to you. It borrows the money from you, and promises to pay you the specific amount it has borrowed from you.
Posted by: Philippe | March 28, 2014 at 11:58 AM
Banks endlessly roll over the debts they owe to depositors, in much the same way that governments endlessly roll over their debts to creditors.
Posted by: Philippe | March 28, 2014 at 12:06 PM
Asymmetric redeemability. I suspect if this is real we would see it in the market: why isn't Goldman Sachs leasing oil tankers and filling them up with dollar bills? Where's the arbitrage? Why isn't someone selling a hedge?
Excess supply of bank money. This is related to the question of what constrains bank expansion, and I suspect David is mostly correct here: bank competition and declining marginal opportunities as their total balance sheet rises as a share of the economy.
Posted by: jt26 | March 28, 2014 at 12:35 PM
Asymmetric redeemability: If I hold a commercial bank deposit the commercial bank owes me central bank money. If I hold central bank money the central bank does not owe me a commercial bank deposit.
(I'm talking about the legal fasts of the real world, not a theoretical world btw)
Posted by: Philippe | March 28, 2014 at 12:59 PM
Holding the price level constant, an excess supply of money equalizes the interest rate on money and non-money, i.e. makes money unprofitable. A central bank could choose to issue money unprofitably, or it could reduce the quantity to maintain a desired interest spread.
Posted by: Max | March 28, 2014 at 12:59 PM
Nick,
banks expand the supply of commercial bank money (CBM) by making loans, which suggests that if there was an excess supply of CBM, it could simply be used to repay loans - thus reducing the amount of CBM. Why would you want the loan but not the money?
Posted by: Philippe | March 28, 2014 at 01:22 PM
Philippe: "Asymmetric redeemability: If I hold a commercial bank deposit the commercial bank owes me central bank money. If I hold central bank money the central bank does not owe me a commercial bank deposit."
Yep.
" Why would you want the loan but not the money?"
Because you want to spend the money.
Max: if the price of refrigerators is perfectly flexible, the price of refrigerators will adjust to ensure the rate of return on holding a fridge is equal to the rate of return on holding money, including convenience yields on both assets.
Posted by: Nick Rowe | March 28, 2014 at 01:42 PM
"If I hold a commercial bank deposit the commercial bank owes me central bank money. If I hold central bank money the central bank does not owe me a commercial bank deposit."
That's a good way of saying it, but I wouldn't agree that it indicates asymmetric redeemability.
The relevant symmetry or asymmetry has to do with the exchange of the two liabilities - central bank money and commercial bank money. That relationship and exchange potential is symmetric, as I described above.
In fact, if you go to the Bank of Montreal with currency, you will receive if desired a Bank of Montreal deposit - and that is ensured mostly by the fact that the Bank of Montreal will receive reserve credit from the central bank in exchange for your currency. That assumes also that the Bank of Montreal will accept you as a depositing customer, but to think of that as an impediment is a pretty far fetched objection, IMO.
What is asymmetric is the agency relationship the commercial bank executes in the two different types of redemption transactions. This operative asymmetry is inherent in the institutional structure and agency relationships of a commercial banking system revolving around a central bank. But it is not an asymmetry in the nature of the exchange of deposits for central bank money and vice versa. The institutional agency asymmetry does not interfere with financial instrument exchange symmetry.
Posted by: JKH | March 28, 2014 at 01:47 PM
comment in spam?
Posted by: JKH | March 28, 2014 at 01:48 PM
Nick, what's the difference between "perfectly flexible" and "perfectly elastic?"
Posted by: Tom Brown | March 28, 2014 at 01:55 PM
i.e.
the relevant interest is in the ability to be able to swap Bank of Canada liabilities for BMO liabilities - in both directions - symmetrically
its not in the physical location where you do the swapping, or in the perceived asymmetry in doing that in one location but not the other
Posted by: JKH | March 28, 2014 at 01:59 PM
i.e. the symmetry in the redeemability of the financial instruments - which is the issue - is unaffected by the asymmetry of either the agency relationship for the redemption or the redemption location
Posted by: JKH | March 28, 2014 at 02:05 PM
JKH,
You can swap BoC liabilities for BMO liabilities, but not with the BoC. Whereas you can swap BMO liabilities for BoC liabilities with BMO. The point is that BoC liabilities can only be passed round amongst other parties - BoC is never obliged to deliver a claim on someone else, like a commercial bank has to.
Posted by: Nick Edmonds | March 28, 2014 at 02:16 PM
JKH,
If you deposit central bank money at a commercial bank, technically you're not redeeming the central bank money for a deposit - you're lending the central bank money to the commercial bank (even if in practice you never actually demand repayment, and just roll over the loan).
There may be rules stipulating that certain types of banks have to accept deposits from anyone - I don't know. But even if that is the case, people are not legally required to become bankers. Nonetheless if you choose to become a banker or to set up a certain type of bank, you might be legally required to accept deposits. Perhaps you know more about the specific rules...
Posted by: Philippe | March 28, 2014 at 02:37 PM
Nick E.,
My point is that the asymmetry of the agency relationship in executing the swap and the asymmetry of the location in executing the swap is secondary to the substance of the swap - which is the symmetry of redeemability of the instrument itself - that must be the central concern - who cares whether the swap is done at BMO or at the North Pole?
And the Bank of Canada is obligated to deliver reserve balances in exchange for currency and vice versa - that's the central symmetry that guarantees the peripheral symmetry of the BMO deposit/currency redeemability feature
Posted by: JKH | March 28, 2014 at 02:44 PM
Nick,
" Why would you want the loan but not the money?
Because you want to spend the money."
If you look at the non-bank sector as a whole, it has to want both the loan and the money. If it didn't want the loan, it would repay the loan thereby extinguishing the money. If it didn't want the money, it wouldn't take the loan, or would repay the loan.
At the individual level, the borrower who takes a loan from a bank wants the loan so that he can spend the money. But he'll also want the money back, so that he can repay the loan (loans have to be repaid).
What he doesn't want is to be stuck with a loan but no money with which to repay it. And repaying the loan deletes the money.
However, whilst the borrower wants the money today so that he can spend it today, he'll only want to get the money back over a period of time, so that he can repay the loan over a period of time. Not sure exactly what that implies for money supply and demand though, as it introduces a time element.
Posted by: Philippe | March 28, 2014 at 02:52 PM
JKH,
I suppose it depends on how you define the word 'redeem'. If it means 'to pay for something', then you could say you buy the bank deposit with the central bank money. But if it means to *repay* something, like a debt, then the term wouldn't be suitable.
Posted by: Philippe | March 28, 2014 at 02:59 PM
Philippe,
I think in the context of Nick's general framing of the issue of asymmetric redeemability, its been about the exchange of a BMO deposit for currency and vice versa
(I think the term redemption is used most generally and usually for asset exchanges - typically redeeming a deposit for a payment - and not (or at least less so) for liability repayment)
Posted by: JKH | March 28, 2014 at 03:07 PM
Nick, good post.
My reading of this post is that it's like all the other posts you've written on the subject. However, for the first time (that I'm aware of) you've included a set of competing assumptions about the nature of deposits vs central bank liabilities, ie. are they perfect substitutes, imperfect substitutes, complements, not substitutes, etc, one's chosen assumption leading to a very different end point. In previous posts, your unstated assumption has always been that deposits and central bank liabilities were substitutes, or at least imperfect substitutes. Even though you're now being more explicit about the menu of assumptions, your preferred selection hasn't changed.
Posted by: JP Koning | March 28, 2014 at 03:14 PM
Set up an alternate clearing system at the North Pole.
Walk into the igloo with a BMO deposit receipt and ask for currency in exchange.
You’ll get it.
Walk in with currency and ask for a BMO deposit receipt.
You’ll get it.
Symmetry
(I wonder if anybody recognizes this as the reverse of that old Stephen Leacock short story.)
In the real world, the commercial banking system acts as a distribution system for central bank currency.
The system is set up as a hierarchy – central bank at the top, commercial banks, then customers, etc.
Hierarchies as distribution systems are inherently asymmetric in structure – the distribution starts at a focal point and spread out.
That’s the asymmetry here.
But the two-way redemption feature as it applies to the two financial instruments in question is symmetric.
Stephen Leacock probably couldn't have created a humorous story of banking without that symmetry.
http://en.wikipedia.org/wiki/Stephen_Leacock
Posted by: JKH | March 28, 2014 at 03:24 PM
http://www.online-literature.com/stephen-leacock/literary-lapses/1/
Posted by: JKH | March 28, 2014 at 03:34 PM
JKH,
"I think the term redemption is used most generally and usually for asset exchanges".
Ok, but when a bank issues a deposit it's not really an "asset exchange". The bank is issuing a liability in exchange for an asset.
Posted by: Philippe | March 28, 2014 at 03:54 PM
Next time I borrow money from a bank I'm going to tell them they're redeeming the money for my loan.
Posted by: Philippe | March 28, 2014 at 04:53 PM
JP: thanks!
Yep, in the past I was never clear in my own mind about whether I was implicitly assuming perfect or imperfect substitutibility.
(Weird thing is: in the old-fashioned treatment, with required reserves, I think reserves and deposits are complements! Not mentally clear on that yet.)
Posted by: Nick Rowe | March 28, 2014 at 06:09 PM
Nick:" suppose BoC promised to convert BMO money into BoC money at par, regardless. So the BoC is the one that pegs the exchange rate. If I were in charge of BMO, I would then print BMO money like crazy, to make lots of profits, and if my customers wanted BoC money I would just get more from the BoC and give it to them."
Which is what happened with the "Affaires des piastres" during the Indochina War. It was even said that it was a reason why the war lasted so long.
https://fr.wikipedia.org/wiki/Affaire_des_piastres ( a less complete text in english available by clicking the english button in the language menu on the left of the page)
I am not a monetary guy but at least I'll play the historian...
Posted by: Jacques René Giguère | March 28, 2014 at 07:07 PM
Nick, O/T: This has to do with Beckworth's latest post on endogenous money. I left a comment on pragcap, and (it's pretty looking with block quotes and all), discussing how your use of the term "exogenous" might differ somewhat from Beckworth's (based on a previous convo between you and I). Maybe you could take a glance and make sure I'm not misrepresenting you:
http://pragcap.com/crickets/comment-page-1#comment-171571
Posted by: Tom Brown | March 28, 2014 at 07:35 PM
Nick, can we use the hypothetical world you established in this post of yours to also examine reflux?
http://worthwhile.typepad.com/worthwhile_canadian_initi/2014/03/the-sense-in-which-the-stock-of-money-is-supply-determined.html
Recall there's just one bank (the CB) and all transactions are in cash (no deposits), and the CB determined a perfectly interest-elastic demand curve for debt at an exogenously fixed rate of interest. Were you already assuming that people could buy their own debt back at par (thus destroying money)? What if we open a second reflux channel and allow (force) the CB to sell the debt to other parties as well, perhaps putting it up for auction? Does this kind of reflux change anything? Is there something more we can learn in regards to reflux from that old example?
Posted by: Tom Brown | March 29, 2014 at 03:11 AM
"It's asymmetric redeemability. This means there cannot be an excess supply of beta money in terms of alpha money."
You didn't take into account other asymmetries such as efficiencies as a means of payment. Commercial bank deposits through electronic payments are a more efficient means of payment than currency. Therefore there can be an excess supply of base relative to deposits IMO.
Posted by: dannyb2b | March 29, 2014 at 03:39 AM
After reading Glasner, I've become (even more) convinced that much of these controversies result from treating results or identities as mechanisms. Examples being things like the money multiplier and national accounts identitities.
For instance saying "suppose the quantity of money doubled" without saying how this came about, isn't sufficient to draw any useful conclusions. So the story about everyone waking up one day having twice as much money doesn't shed that much light on what happens when the Fed conducts open market operations.
These tautologies are quite different from descriptive models like the Lavoie and Godley equations, which are detailed accounting flows, and these are in turn different from NK Euler equations which are exact predictive models, but need not be stock and flow correct.
Posted by: Peter N | March 29, 2014 at 05:01 AM
JKH,
I think the point about the asymmetry is that it is mainly about who is committing to exchange instruments and who is not. Commercial banks have to carry out certain exchanges that the central bank does not.
In particular, the central bank does not have to do exchanges that involve credit asset / debit liability, which is what commercial banks do when they deliver currency in repayment of deposits. This means that commercial banks face the possibility of their balance sheets shrinking to zero, which affects how they can set their rates. The central bank does not face this constraint and so can set its rate at whatever it likes.
I'm not sure whether asymmetric redeemability is the best term for this, but I do think this is an essential feature of what makes the central bank special and I think this is equivalent to what Nick describes asymmetric redeemability as being.
Posted by: Nick Edmonds | March 29, 2014 at 06:21 AM
"Yep, in the past I was never clear in my own mind about whether I was implicitly assuming perfect or imperfect substitutability."
"It seems reasonable to assume that commercial bank money is an imperfect substitute for central bank money."
So your refrigerator analogy no longer holds, at least with respect to bank issued deposits. You used to say that if people don't want to hold more stocks of furniture, it would be impossible for suppliers of furniture to increase the stock in public hands, but that with money it was different. This post puts forward a different set of assumptions. Now banks are like furniture suppliers, they need to make their "product" slightly more attractive in order to get the market to hold it. So you've moved from perfect to imperfect substitutability?
Posted by: JP Koning | March 29, 2014 at 07:25 AM
JP: imagine many producers of fridges. The beta producers promise to exchange their fridges for alpha fridges at par.
Any producer, whether alpha or beta, could not even get people to accept more fridges, unless he made them more attractive to hold. Even if all producers try to expand together, they will fail, unless they make fridges more attractive to hold.
A beta producer of money can get people to accept more money, but they will not hold it unless he makes it more attractive to hold, or unless all money is perfect substitutes. If all producers expand together, they will "force" people to hold more money.
Posted by: Nick Rowe | March 29, 2014 at 08:29 AM
Peter N: "NK Euler equations which are exact predictive models, but need not be stock and flow correct."
Show me an NK model which is stock and flow incorrect.
"Were you already assuming that people could buy their own debt back at par (thus destroying money)?"
They could, but they may not want to. If the bank lets people sell back their money for apples, there cannot be an excess supply of money in terms of apples, but there can be an excess supply of money in terms of all other goods.
Posted by: Nick Rowe | March 29, 2014 at 08:34 AM
Nick, thanks for you answer. Does your answer there speak to your seeming disagreement with David Glasner over reflux?
http://uneasymoney.com/2014/03/27/the-uselessness-of-the-money-multiplier-as-brilliantly-elucidated-by-nick-rowe/#comment-77508
You guys are seemingly close on this last I checked... would he agree with your statement here about the apple market (as if the bank were buying and selling apples)? I'm having a hard time figuring out exactly how you two disagree or if you still do.
Posted by: Tom Brown | March 29, 2014 at 10:11 AM
Nick (or anyone!), you write about your 3rd case (not substitutes and not complements):
"the Fed could not create inflation by increasing the supply of beta money and making it more attractive to hold."
I assume you're not implicitly restricting the inflation to either Canada or the US here?
And it's a case of not substitutes or complements because only $alpha used in Canada and only $beta used in US?
But still you're imagining some kind or arbitrage or trade across the border. Can somebody walk me through this? Say the Fed increases the supply of beta money and makes it attractive to hold... now what?
Also on the 4th example: the "complement" one. I assume that can't be too important, since you only put it there "for completeness" ... but still, why bring it up at all? I'm not sure I get what you mean by complement... not this: $beta = 1 - $alpha, right?
Posted by: Tom Brown | March 29, 2014 at 10:36 AM
JP,
Thanks for interpreting further. That’s a clear and interesting explanation.
So as I understand it, the scenario of interest is the conversion of bank deposits into currency – a flight to quality from commercial bank deposits to central bank currency.
That shrinks the banking system balance sheet with a lower bound of zero – sort of.
In the other direction, the scenario of interest is the conversion of currency into bank deposits – a flight to commercial banking deposits from central bank currency.
That expands the banking system balance sheet with an upper bound of an addition equal to the size of the central bank balance sheet – sort of.
I’m not finding any disastrous asymmetry there – other than the comparative sizes of the CB balance sheet and the commercial system balance sheet.
There’s an efflux of currency where the commercial system balance sheet gets absorbed by the central bank and a reflux of currency where the central bank balance sheet gets absorbed by the commercial system (in the sense that the initial funding interface between the two is excess reserves)
In fact, that’s starting to sound a bit symmetric to me.
Now – entirely separate issue - I would say the risk comparison between those two scenarios is different.
But that’s risk analysis.
It has nothing to do with a concept of “asymmetric redeemability” that I can detect.
So my point again is that Nick is exploring some aspect here that is different from “asymmetric redeemability”.
Also, when you say this:
“In particular, the central bank does not have to do exchanges that involve credit asset / debit liability, which is what commercial banks do when they deliver currency in repayment of deposits.”
OK. But is that Nick’s point? That the central bank doesn’t deliver an asset (like gold) from the asset side of its balance sheet? I haven’t seen that anywhere. I think that’s a different issue again. I don’t see where Nick has specified his differentiation between symmetric and asymmetric on the basis of the central bank delivering fiat currency (a central bank liability) rather than something like gold (a central bank asset). If that’s Nick’s point, wouldn’t it have been easier to say that – i.e. that by contrast, symmetric redemption exists in the case of gold? And would Nick say this? I doubt it.
Nick does say this:
“Commercial banks are typically beta banks, and central banks are typically alpha banks. Beta banks promise to convert their money into the money of alpha banks at a fixed exchange rate. Alpha banks make no such promise the other way. It's asymmetric redeemability.”
JP, I see nothing there about the critical role that delivering a central bank liability rather than a central bank asset is presumed to play in your interpretation of Nick’s declaration of asymmetry.
But if that’s what Nick is saying, then maybe I’ll have to reconsider.
So question for Nick is – if I can exchange gold for my BMO deposit – at my BMO branch - either direction – does that constitute symmetric redeemability according to your model – and is that special criterion consistent with your criterion for asymmetric redeemability in the case of fiat money?
One more time (sigh!), regarding my own view of symmetry and asymmetry:
If I have a deposit account at BMO, there is no way that BMO will not credit my account when I deposit currency.
(If I don’t have a deposit account at BMO, and if I can’t get any other Canadian bank to allow me to open up a deposit account, then I have a much bigger problem with my banking arrangements, and one which has nothing to do with the general issue of symmetry or asymmetry.)
Yet that is the particular transaction that seems to be the bone of contention in the question of symmetry or asymmetry. As far as I can tell, the argument for asymmetry seems to be that it is BMO rather than the Bank of Canada that is prepared to credit my BMO account.
I find that argument to be irrelevant to the substance of the issue. I am interested in being able to exchange my BMO balances for central bank paper – in either direction. And there’s nothing preventing me from doing that.
As I said earlier, I don’t care who the agent is for the transaction. I only care about being able to exchange the paper. The fact that I don’t have a deposit account at the central bank where I might be able to do the same thing in exchange is a mere institutional quirk. I would have thought Nick would recognize that sort of thing as being secondary.
The issue is BMO paper (i.e. a deposit receipt or printout) and Bank of Canada paper.
What am I allowed to redeem and in which direction?
The answer seems simple to me.
Any other complication that might impede that transaction is an issue that is beyond the question of symmetric or asymmetric redeemability, IMO. In that sense, I agree with you that whatever it is that Nick wants to draw out, it may not be best described as “asymmetric redeemability”. My problem is I really don’t know what that precise issue of interest is. I’m sure there is one, but it’s being blurred in my view by this particular construction of choice.
P.S.
Nick, I’m pursuing this because I find it interesting – not because I’m an ankle biter (I hope)
Posted by: JKH | March 29, 2014 at 10:44 AM
Suppose you're a bank and people can use the money you issue either 1) to swap for the money of another bank or 2) swap for other goods or 3) hold it
You can vary both the amount of money you issue and the interest you pay people to hold your money.
Suppose you want to keep the exchange rate of your money against the other money fixed, while increasing the qty you issue: You would do this by increasing the interest rate you pay on your money.
Suppose you want to keep the exchange rate of your money against goods fixed, while increasing the qty you issue: You would do this by increasing the interest rate you pay on your money.
Based on this: are there not at last some situations where you can increase both the qty of money and interest rates and keep your exchange rate against both the other currency and goods constant ? You do not necessarily create an excess demand for money because even though the combined qty of your money and the other money has increased people are holding more of it and not spending it because of the higher interest rate.
Posted by: The Market Fiscalist | March 29, 2014 at 11:43 AM
**"excess supply of money" not "excess demand"
Posted by: The Market Fiscalist | March 29, 2014 at 11:50 AM
JKH,
commercial banks are ultimately legally obliged to redeem bank deposits for central bank money (if that's what the depositor/creditor demands), but the central bank is not legally obliged to redeem central bank money for commercial bank deposits.
This is an asymmetric relationship due to the fact that commercial bank deposits are bank debts denominated in central bank money, whereas central bank money is not a central bank debt denominated in commercial bank deposits.
Commercial banks aren't obliged to accept central bank money in exchange for deposits, nor are central banks obliged to accept commercial bank deposits in exchange for central bank money. This is symmetrical because people aren't legally obliged to become indebted.
The fact that as a customer of a bank you can exchange cash for deposits both ways easily (usually), unless the bank is illiquid or insolvent, doesn't mean that there is a symmetric relationship.
If the central bank did promise to redeem its liabilities for something like gold, that would not make the relationship between central bank money and commercial bank deposits symmetrical, because the central bank would still not be obliged to redeem its money for commercial bank deposits.
Posted by: Philippe | March 29, 2014 at 11:52 AM
Market Fiscalist,
"Suppose you want to keep the exchange rate of your money against the other money fixed, while increasing the qty you issue: You would do this by increasing the interest rate you pay on your money."
It seems to me that if you promise to redeem your money for something else, or to convert it at a fixed exchange rate, just raising the interest you pay could lead to insolvency unless your revenue also increases. If you issue purely 'fiat' money which can not be redeemed or exchanged at a fixed rate, then raising the interest you pay could potentially be inflationary unless you increase your revenue.
Posted by: Philippe | March 29, 2014 at 12:02 PM
"Peter N: "NK Euler equations which are exact predictive models, but need not be stock and flow correct."
Show me an NK model which is stock and flow incorrect."
There is nothing I know of that forces one to be incorrect. They certainly have things like budget constraints and no-Ponzi conditions, that prevent pathological results, and, IIR assuming rational expectations produces a kind of formal reconciliation, but the result may be vacuous. An example would be accounting for investment in a representative agent model. The fine details are a bit above my pay grade so I'll defer to an expert:
http://www.ipc-undp.org/publications/srp/Notes%20on%20the%20Stock-Flow%20Consistent%20Approach%20to%20Macroeconomic%20Modeling.pdf
I suspect, however, that you've seen this stuff before.
Posted by: Peter N | March 29, 2014 at 12:17 PM
Phillipe,
Yes I agree. I was assuming that the bank could ignore profit and loss, and other market considerations, and was just interested in the pure exchange rate (against goods and other currency) situation.
If the bank was profit seeking then there would still be a combination of money it creates, interest paid, price level and exchange rate that would maximize profits - and it would adjust its supply of money and interest rate paid to find this level. (Of course the exchange rate may be fixed at 1:1 by law, and the other bank may keep adjusting its money supply so that this bank finds it also has little control over interest rates too).
Posted by: The Market Fiscalist | March 29, 2014 at 12:18 PM
I've been reading a lot of stuff I agree with in your recent writing. Perhaps you should be worried?!
Posted by: Peter N | March 29, 2014 at 12:21 PM
JKH, imagine a new banking system starting fresh, with an all cash central bank (no electronic reserve deposits). Say the central bank buys exactly $1 in assets, and it announced that's the last OMO it will perform for the next 30 years. Commercial banks are going to be very hesitant to expand deposits beyond $1 total knowing that they must guarantee to redeem their deposits on demand for cash.
The central bank has no such concerns. It can change it's mind tomorrow and buy another $1 if it wants. It has no limit on the growth of it's liabilities imposed on it by the commercial banks. The commercial banks live in a totally different reality.
Posted by: Tom Brown | March 29, 2014 at 02:47 PM
Tom Brown,
"it announced that's the last OMO it will perform for the next 30 years"
interesting forward guidance strategy
et tu, Brute?
:)
Posted by: JKH | March 29, 2014 at 03:33 PM
As far as I am aware, there is no existing legal obligation according to which the holder of a central bank's physical currency can bring a central bank-issued currency note to the central bank and demand the central bank convert it into either a deposit balance at a commercial bank or into some other kind of commercial bank obligation; nor can a commercial bank that is the holder of a reserve deposit balance at the central bank demand that the central bank convert that balance into either a deposit balance at a commercial bank or into some other kind of commercial bank obligation.
However, a commercial bank is legally committed to exchange its own deposit account obligations into central bank obligations on demand. So it seems to me that Nick's claim of asymmetric redeemability is obviously correct.
Posted by: Dan Kervick | March 29, 2014 at 04:04 PM
Peter N: 20 years ago I saw one published micro-founded NK model that was "stock-flow inconsistent". The authors had forgotten to include firms' profits in the agents' budget constraint, and that omission caused their very weird (keynesian) results. It was a simple mistake, and any economist, who figured out that mistake, would say it was a problem. The whole point of microfoundations and "DSGE" models is to (try to) make sure everything in the model is internally consistent both in each time-period and across time-periods. (Though internal consistency, of course, is not the only thing one hopes for in a good model, and some of us think that maybe micro-founded models pay *too much* attention to internal consistency.)
That is one NK paper with an accounting mistake. If that mistake had been spotted by the referees, they would probably have insisted it be fixed before the paper was published. There may be others, but I haven't spotted any.
Posted by: Nick Rowe | March 29, 2014 at 04:40 PM
Tom:
Elasticity
Substitutes and complements.
JKH: carry on. No problems.
Suppose (as under a sort of Bretton Woods gold standard): BMO fixes its exchange rate to BoC paper, and BoC fixes its paper to Fed paper, and the Fed fixes to gold. (Except in the real world it wasn't exactly like that, because normal people couldn't go to the Fed and demand gold for Fed currency, IIRC.)
BMO is beta to BoC's alpha, but the BoC is in turn beta to Fed's super-alpha, and the Fed is in turn beta to the gold miner's mega-alpha.
Posted by: Nick Rowe | March 29, 2014 at 05:01 PM
"If alpha and beta money were perfect substitutes for each other, people would be indifferent about the proportions of alpha to beta monies they held."
I'm struggling with this concept. If I'm a beta bank this would mean that when I issue loans in beta money and people spend that money it will never come back to me to be converted to alpha money. Its like I have a printing press that can produce perfect counterfeits of alpha money. If I keep printing then of course this will be inflationary (unless countered by less printing by the alpha bank). To prevent this scenario alpha banks will always have an incentive to prevent beta money being perfect substitutes for its money.
But as soon as there is the smallest difference between the beta currency and the alpha currency then as soon as people see signs of over issue of the beta currency they will move to "reflux" it back into alpha money. So the beta-issuing bank can never generate inflation beyond that which simply mirrors alpha money inflation.
I can't think of any scenarios outside of counterfeiting where alpha and beta money would be perfect substitutes so I think David Glasner is right on this point.
Posted by: The Market Fiscalist | March 29, 2014 at 05:03 PM
TMF: "But as soon as there is the smallest difference between the beta currency and the alpha currency then as soon as people see signs of over issue of the beta currency they will move to "reflux" it back into alpha money. So the beta-issuing bank can never generate inflation beyond that which simply mirrors alpha money inflation."
No. The individual beta bank can just pay a tiny amount of interest (in one form or another) on its money, to offset people's desire to convert it into alpha money. But they still want to convert both monies into other goods, so that creates inflation.
Posted by: Nick Rowe | March 29, 2014 at 05:38 PM
Nick,
I have no fundamental problem with the alpha/beta paradigm in the sense of being open to some kind of focused and limited interpretation of it. It can be separated from the idea of deposit insurance for example. Deposit insurance can be abandoned provided that sufficient interest is paid on deposits (if necessary) in order to allow for perceived risk to deposit value after taking into account the first loss position of equity and other forms of capital. Authorities could conceivably accommodate a partial flight to currency (perceived) or other banks before deciding to shut a bank down. In any case, what is being protected is the value of the aggregate quantity of dollars in a deposit - not the value of each of those dollars.
The question of whether losses can wipe out deposits in individual banks is really separate from what gives a currency its value. In that sense, I think of alpha/beta as being at the same general level of analysis as Chartalism – the way(s) in which a currency holds its value. Which is separate from the question of whether losses in banks can wipe out deposits. No real problem with the BMO/Fed/gold system alpha/beta paradigm as described.
But I view all of that as quite separate to the question I’m raising on symmetric or asymmetric redeemability as described.
Posted by: JKH | March 29, 2014 at 06:15 PM
"The individual beta bank can just pay a tiny amount of interest (in one form or another) on its money, to offset people's desire to convert it into alpha money. But they still want to convert both monies into other goods, so that creates inflation."
OK. I think I see it.
There are 2 obvious reason to switch between alpha and beta money
1. The inflation rate is higher in one than the other
2. The interest rate is higher in one than the other.
In addition banks can avoid inflation by paying higher interest rates on their money.
Beta banks increase the supply of their money. They know that if this increases beta-inflation it will cause reflux so they also increase the rate of interest. This avoids inflation and maintains the exchange rate.
However the rate of interest is now higher on beta than alpha and causes people to swap from alpha to beta money. This causes beta money to appreciate in value. The beta bank can then safely increase the money supply (and the inflation rate) until the old exchange rate is restored. Both alpha and beta prices are now higher.
The problem I see with this is: Outside of a simple model the [beta] banks will have to set the interest rate to a market clearing and profit-maximizing rate. They can't just increase the money supply and the interest rate without moving out of profit-maximizing equilibrium.
So: I now think you are right in theory but David is in practice :)
[edited to fix typo NR]
Posted by: The Market Fiscalist | March 29, 2014 at 06:54 PM
"Outside of a simple model the alpha banks...." should be "Outside of a simple model the beta banks...."
Posted by: The Market Fiscalist | March 29, 2014 at 06:56 PM
Nick,
I find it very useful to extend this alfa-beta discussion by considering the role of bank capital.
Beta market share is limited by capital. When all beta banks lose capital (e.g. housing bubble losses), they lose market share. Then it makes sense for alfa bank to step in and do QE, thereby expanding alpha share even further.
Alfa bank benefits from assymetric redeemability, thus alfa bank needs no capital. Except when alfa bank has to follow super-alpha bank. Most central banks have their super-alpha banks (for example the Fed targets a narrow range for 5y5y forward inflation breakeven spread ). Thus most central banks are banks, and they need capital.
Posted by: Vaidas Urba | March 29, 2014 at 07:03 PM
JKH, "et tu Brute?" Haha... Sorry to pile on there, but I'm still not getting what you're getting at I guess. In my simple example, isn't it just a case of asymmetric redeemability? The CB can close up shop for 30 years... they don't have to worry about redeeming jack at that point. It's up the to the commercial banks to sort it out. If the CB wants to surprise everyone and do something different in an hour, then it's up to everyone else to adapt. The commercial banks have to redeem their deposits for cash, and they may choose to accept cash for their deposits... so say they all do. Then that part is fully symmetric: cash for deposits and deposits for cash. The asymmetric part is who's doing the redeeming: the CB does what it wants in this idealized world w/o regard to the commercial banks and their constraints.
Now in reality the CB may be a puppet alpha bank, beholden to the bank representatives that sit on its board, who decide that inflation above what they were expecting is VERY bad for creditors, so that's NOT EVER going to happen.... in fact, inflation might just continue to slide incrementally lower and lower than expected, because that helps the creditors. After all the CB was put in place to serve their interests. Sorry, I'll take my tinfoil cap off now. :D
Posted by: Tom Brown | March 29, 2014 at 08:17 PM
TMF: "The problem I see with this is: Outside of a simple model the [beta] banks will have to set the interest rate to a market clearing and profit-maximizing rate. They can't just increase the money supply and the interest rate without moving out of profit-maximizing equilibrium."
Understood. So let me rephrase my assertion: if something (like an increase in beta bank capital, following Vaidas) causes that profit-maximising equilibrium to change, so that beta banks find it profitable to increase their supply of beta money, and raise interest rates on beta money so that people do not wish to convert the extra beta money into alpha money, that will create an excess supply of both alpha and beta money against all other goods, for a given stock of alpha money.
JKH: suppose the actual ratio of beta/alpha money is higher than people want to hold. The alpha bank may *choose* to increase the stock of alpha money to reduce that ratio to the desired ratio, if it thinks it needs to do so to keep hitting its monetary policy target. But, if the alpha bank does not choose to do that, the beta banks will be *forced* to reduce the stock of beta money, or else make beta money more attractive to hold.
Posted by: Nick Rowe | March 30, 2014 at 04:19 AM
Tom,
The normal deal is that the economy needs more central bank notes as it grows. Bank customers get them from their banks. No 30 year shut down allowed.
(BTW, you’re off my short list for next Fed chair.)
“Then that part is fully symmetric: cash for deposits and deposits for cash. The asymmetric part is who's doing the redeeming.”
And my point is we shouldn’t care who is doing the redeeming (i.e. who the distribution agent is). If I can go to BMO and exchange either way, I don’t need to go the central bank to do the same thing. The absence of operational duplication is a logistical asymmetry that poses zero economic cost to the system. It is moot. It is an irrelevant asymmetry. It is a dead parrot asymmetry. It doesn’t matter.
Conversely, the available symmetry of exchange at BMO is meaningful. Particularly if Chairman Tom Brown isn’t allowed to shut it down for 30 years.
Posted by: JKH | March 30, 2014 at 06:00 AM
Nick,
ratio beta/alpha high
then:
alpha bank can ease
or beta banks can tighten
ratio beta/alpha low:
then:
alpha bank can tighten
or beta banks can ease
symmetric
Maybe the first is what happens at the lowly operational level of bank reserves as described by endogenous money adherents – the CB supplies reserves after deposits have been created by loans
The second example at the same level could be the reverse of that (alpha withdrawing reserves) or a sort of QE (no withdrawal, encouraging beta expansion)
Posted by: JKH | March 30, 2014 at 06:13 AM
Chasing symmetry:
http://en.wikipedia.org/wiki/Turtles_all_the_way_down
Posted by: JKH | March 30, 2014 at 06:17 AM
@Nick Rowe, at 04:19 AM
"beta banks find it profitable to increase their supply of beta money, and raise interest rates on beta money so that people do not wish to convert the extra beta money into alpha money"
But if you look at (1yr CD rates)-(1y Tsys), the spread has been between -2 to 1.5 % (2000-2013); -2% in 2006 with the economy booming. The currency component of M1 isn't correlated at all. Neither is the federal deficit (issuance of Tsys). On the other hand, except in times of crisis, AA financial commercial paper is very tightly correlated with fed funds. I'm not sure if there is a single coherent story of "supply and demand" for money.
At least for commercial paper funding, the story looks like the banks just pay fed funds + 0.1%, and then add their net interest margin on top.
One CD story, looks like customers need relatively rapid liquidity and take whatever the banks feel like offering. This makes sense because of precautionary waiting periods for financial transfers, check clearing etc. The frictions add additional liquidity premium.
Another CD story could be, that a rapidly growing economy with rapid lending increase, requires more liquidity services, so it may be as much about flow as stock (2006). M1 is tiny compared to the net assets in the economy, so small changes in asset stock or stock turnover can create large flows. So is there an excess supply of bank money, or a strange dynamic between bank lending, liquidity services and asset transformation velocity?
Posted by: jt | March 30, 2014 at 10:29 AM
JKH
"..my point is we shouldn’t care who is doing the redeeming."
But that is the key distinction here, because it's not about whether you can convert either way at par, but rather about which entities are obliged to deploy their balance sheet in maintaining the par conversion rate (ultimately to the point where they are insolvent).
Posted by: Nick Edmonds | March 30, 2014 at 11:02 AM
Nick E.,
Both entities have to deploy their balance sheets, of course.
But since when did insolvency become the test of asymmetry of redemption?
Did Nick R. specify that as the criterion?
Didn't see it.
But if so, call it solvency asymmetry - not redemption asymmetry.
Redemption symmetry is a liquidity issue - not a solvency issue.
All banks are still able to order currency prior to being shut down.
Solvency asymmetry is a much larger issue than redemption risk.
Posted by: JKH | March 30, 2014 at 11:21 AM
In addition to that, the "bone of contention" in the asymmetry question is not whether BMO deposits can be converted to Bank of Canada notes, but vice versa. That is what Nick has said (see his first paragraph here and other posts.) That is what I have responded to so far.
That defining concern is upside down and inconsistent with respect to the suggestion that insolvency risk is the defining criterion.
Posted by: JKH | March 30, 2014 at 11:30 AM
In other words, Nick R.'s defining criterion for asymmetric redemption is the opposite of Nick E.'s.
Nick R.'s ultimate test equates to whether the CB will actually shrink its balance sheet to accommodate BMO expansion. And there's nothing to prevent that if that's what BMO customers want.
I dealt with that earlier above.
Posted by: JKH | March 30, 2014 at 11:42 AM
Maybe the mention of the limit of insolvency was misleading, because the point is not about how far they have to deploy their balance sheet, but whether they have to do so at all. The central bank is not obliged to use its balance sheet to maintain par conversion, at all.
I agree it's more relevant as a question about whether BoC notes can be converted in BMO notes than the other way around. But it's not about whether you as an individual can do that, but about whether the system as a whole can do it. In other words can the BoC notes be converted out of existence. And as the only way to do that is by converting them with BoC itself, it comes down to whether BoC is obliged to do the conversion.
Again, I'm only saying what I think the asymmetry is and I didn't come up with the term. But I think this is equivalent to what Nick is describing.
Posted by: Nick Edmonds | March 30, 2014 at 11:51 AM
JKH, I think you are arguing for my "puppet" alpha banks. Your alpha banks are not true alphas: they are beholden to the banks, or their customers, etc. They don't get to be tyrants that can do exactly as they please, including 30 year vacations.
Posted by: Tom Brown | March 30, 2014 at 11:59 AM
... my puppet alphas can be beholden to the alpha's own plan too. But sometimes that plan serves the purposes of others, and sometimes it's capricious and arbitrary (like my plan to target base money at $1 for 30 years... a plan though! ... like any other!).
Posted by: Tom Brown | March 30, 2014 at 12:04 PM
... if the criteria for success of the alpha is they adopted a target, stuck with it, and consistently hit the target (minimizing error), then it's hard to beat the MB = $1 target: They can NAIL that target exactly (error = 0 always), and thus justify their long vacation.
Posted by: Tom Brown | March 30, 2014 at 12:07 PM
Nick E.,
"The central bank is not obliged to use its balance sheet to maintain par conversion, at all... it's about whether the system as a whole can do it. In other words can the BoC notes be converted out of existence. And as the only way to do that is by converting them with BoC itself, it comes down to whether BoC is obliged to do the conversion."
The BOC absolutely is obligated to take in notes in exchange for reserves - which is all the banks want in order to offer deposits in exchange for notes. Again, I'm assuming the idea that bank customers can't deal directly with the Bank of Canada is moot - because they can get everything they want by dealing with their commercial bank of choice.
Maybe one of the differences in views here is that I'm assuming that the rules are in place - i.e. that operations continue as is.
When you say "not obliged" - well, the CB is not obliged to do anything - to the degree that it's part of the state complex that can change any of the rules at a whim.
I do interpret Nick's interpretation as more operational than that.
Posted by: JKH | March 30, 2014 at 12:29 PM
Nick,
"20 years ago I saw one published micro-founded NK model that was "stock-flow inconsistent".
This is an example of my original point about the problems that arise from using the same terms to mean different things. Clearly you and Godley have very different ideas of what it means to be stock-flow consistent.
"Foley (1975) proved that, under the assumption of “perfect foresight on average” , the distinction between stock and flow equilibria in asset markets is non-existent. In other words, under the assumption of rational expectations there’s no logical problem in phrasing macroeconomic models just in terms of flows (or stocks), since a flow (stock) equilibrium would necessarily imply a stock (flow) equilibrium as well."
So you're correct in your own terms since almost all NK models assume "perfect foresight on average", and even this is only a sufficient condition, not AFAIK a necessary one.
But if you believe that
"As Godley and Cripps (1983, p.18) eloquently put it, 'the fact that money stocks and flows must satisfy accounting identities in individual budgets and in the economy as a whole provides a fundamental law of macroeconomics analogous to the principle of conservation in physics.'"
then you need the equivalent of quadruple entry bookkeeping and sufficient entities for meaningful flows (like banks, for instance, to recycle savings). This leads to a different conclusion. I was making the distinction between NK style models and Godley style models, and I think Godley's own terms are the appropriate ones for doing this.
This, of course, is all in the Godley-Lavoie book.
http://www.amazon.com/Monetary-Economics-Integrated-Approach-Production/dp/0230301843
Posted by: Peter N | March 30, 2014 at 12:32 PM
JKH,
OK, but when I said about BoC notes being converted out of existence, I didn't mean by conversion into other liabilities of BoC, I meant into liabilities of someone else, e.g. of BMO. i.e. I meant having to do credit asset / debit liability, not credit liability / debit liability.
The fact that the central bank does not have to do this exchange is critical to its status.
Posted by: Nick Edmonds | March 30, 2014 at 12:53 PM
Nick E.,
"The fact that the central bank does not have to do this exchange is critical to its status.
Disagree totally there, Nick.
It's entirely moot in my view.
The public wants BMO deposits in exchange for central bank notes - and vice versa.
Doesn't matter who does the conversion.
What matters is the credit/liquidity characteristic of the instrument and the reason for the conversion - not who is the agent for the conversion.
What seems to be lost within this view of asymmetry is that the commercial banks act as distribution agents and dealers for central bank notes. Indeed, that essential dealer function IS why the flow is symmetric - even though a distribution network is an inherently asymmetric organization structure (for good reason).
So BMO gets the reserves for notes - and the customer gets the BMO deposit for notes.
The central bank has to do that exchange with BMO - and BMO has to do that exchange with the customer - assuming the customer has a deposit account. And if the customer doesn't have a deposit account with any bank, that's a different problem.
And the fact that the CB ends up with reserves is also moot - it just sells assets if it wants to wind down reserves.
Posted by: JKH | March 30, 2014 at 01:37 PM
JKH,
in one of your replies to Nick Rowe above, you wrote:
"...there is no asymmetry in terms of the option to redeem in either direction at any time.
Unless a bank is shut down – then all bets are off."
http://worthwhile.typepad.com/worthwhile_canadian_initi/2014/03/there-can-be-an-excess-supply-of-commercial-bank-money.html?cid=6a00d83451688169e201a51190e296970c#comment-6a00d83451688169e201a51190e296970c
Isn't that the point? Commercial banks can become insolvent if they are no longer able to redeem their deposits for central bank money. But central banks cannot become insolvent due to not being able to redeem their money for commercial bank deposits - because they make no such promise. This is the asymmetry.
Posted by: Philippe | March 30, 2014 at 02:35 PM
So, say the central bank decides to buy bills and it does so by buying them from non-banks and paying currency. Now the public has more CB notes than they want so they take them to BMO and exchange them for BMO deposits. BMO exchanges them with the CB for reserves.
Now, if BMO finds itself with more reserves than it wants, it might like to go along to the CB and exchange them into BMO liabilities (which the CB would have to somehow acquire), which it can then cancel. If it could do this, it could happily swap currency for deposits with the public all day, because it could back it out with the CB. But this is exactly the trade that the CB won't do and it's why the CB can force extra reserves into the system without them flowing straight back. Which is how it can control the price.
Posted by: Nick Edmonds | March 30, 2014 at 02:51 PM
Phillipe,
“Commercial banks can become insolvent if they are no longer able to redeem their deposits for central bank money.”
That’s not quite the same as what I would say, which is:
“Commercial banks can always redeem their deposits for central bank money – unless they are declared insolvent by the central bank.”
As long as a commercial bank is up and running, the CB must respond to orders for currency or redemption of currency.
If the commercial bank is declared insolvent, all bets are off in the sense that nothing is operative – including redemption of deposits for currency.
I agree that the issue of insolvency itself is one of asymmetry (although that does require technical state support for its central bank, but that shouldn’t be considered to be a realistic impediment.) But as I also said, the issue of insolvency is larger and separable from this issue of asymmetric or symmetric redemption.
Central banks DO make the promise to redeem their money for commercial bank deposits – and vice versa - using the commercial banks as agents. Again, it’s not the agent of redemption that is the relevant issue – it is the pieces of paper/deposits that are being exchanged. That’s what the customer cares about – exchanging central bank currency for a BMO deposit and vice versa.
Central banks only stop making that promise when they shut a bank down – which is typically a solvency issue. And that involves a lot more than shutting down currency operations. Solvency is a bigger issue than liquidity.
Posted by: JKH | March 30, 2014 at 03:18 PM
Nick,
I can’t seem to make sense of your example, or where it’s going.
“Now, if BMO finds itself with more reserves than it wants”
At that point, it’s effectively a QE balance sheet situation, achieved through currency injection followed by a currency/reserve conversion.
“BMO liabilities which the CB would have to somehow acquire”
That required acquisition is equivalent to an additional direct QE transaction with BMO – it increases reserves.
Then you seem to be cancelling something which was a prerequisite to its own cancellation.
So no net effect.
That’s what it looks like to me.
If I misunderstand, maybe you could do the accounting in a bit more detail.
Although your example is seeming like a stretch in progress.
Posted by: JKH | March 30, 2014 at 03:40 PM
JKH, you write:
"Central banks DO make the promise to redeem their money for commercial bank deposits"
I thought they just made a promise to exchange electronic CB-deposits (liabilities to the CB) for reserve notes and coins. No?
"As long as a commercial bank is up and running, the CB must respond to orders for currency or redemption of currency."
To the extent that the commercial bank holds CB-deposits, but for commercial bank-liabilities too?
Isn't the CB legally off the hook as long as it agrees to deliver enough paper to back the outstanding electronic CB-deposits, but no more? In the US the Fed could call up the BEP, order up enough paper notes to back all outstanding electronic Fed-deposits, and then never call the Bureau of Engraving and Printing (BEP) again: and they'd be all set. The amount of paper notes sitting at the Fed, waiting to become Fed liabilities is then fixed and has nothing to do with the amount of deposits that commercial banks might create: if the commercial banks create too many and get themselves into trouble, then that's their problem, isn't it (ignoring the regulatory role of the Fed for the moment)?
And if the Fed eventually has unused currency in excess of all outstanding Fed deposits, they can safely burn the excess if they want (assuming again, no further orders placed to the BEP for more paper notes). Forget about coins the moment, but it's basically the same story (with a little different accounting).
Posted by: Tom Brown | March 30, 2014 at 03:58 PM
Nick, do you have a response to David's latest response to you on his blog?
Posted by: Tom Brown | March 30, 2014 at 04:02 PM
JKH,
"Central banks DO make the promise to redeem their money for commercial bank deposits – and vice versa - using the commercial banks as agents."
No they don't. Central banks are not obliged to accept deposits at commercial banks in exchange for their money. A commercial bank can not force the central bank to pay it cash in exchange for a commercial bank deposit.
Posted by: Philippe | March 30, 2014 at 04:11 PM
Tom,
It’s transitive logic.
The CB makes a promise to redeem currency for reserves and BMO makes a promise to redeem currency for deposits. The CB's promise allows the BMO's promise.
Therefore the CB makes a promise in effect to redeem currency for deposits.
The reason for all this is that the commercial banking system is the anointed distribution and dealing agent for central bank currency.
If there’s any asymmetry associated with this issue of currency flow – it’s that – the distribution network system – not the two way flow of redeemability. And the network structure asymmetry is simply due to the existence of a commercial banking system instead of one single (central bank).
I think there are a number of underlying issues of monetary asymmetry here. Solvency and network structure are two of them. But redeemability is not in my view.
Posted by: JKH | March 30, 2014 at 04:12 PM
Phillipe,
What I said in my response to Tom Brown.
Posted by: JKH | March 30, 2014 at 04:15 PM
Phillipe,
You go to your bank with currency.
You have a deposit account with that bank.
The bank refuses to take your currency in exchange for a credit to your deposit account.
That's against the law.
The CB/regulator shuts down your bank.
+ my comment to Tom
= earlier comment
Posted by: JKH | March 30, 2014 at 04:20 PM
but banks are not obliged to accept any and all deposits. They have no legal obligation to enter into a contract with a prospective customer.
However, once they have a legal contract with a depositor, they are obliged to abide by that contract, which includes redeeming deposits for central bank money.
Posted by: Philippe | March 30, 2014 at 04:56 PM
JKH, you write:
"The CB makes a promise to redeem currency for reserves and BMO makes a promise to redeem currency for deposits. The CB's promise allows the BMO's promise."
I agree the CB could choose to make that promise, and practically they may be compelled to make that promise (due to banker's sitting on their governing board), but I don't see why they necessarily legally (by the book) have to make that promise. But then I'm absolutely no expert...
I think this gets to one core issue separating PKEers from MMists: PKEers assume the CB is fundamentally accommodative (at some core level) to the banking system, while MMists do not. What do you think?
Could it be that by the letter of the law that MMists are generally correct, but practically speaking reality may lean a bit more towards the PKE position?
Posted by: Tom Brown | March 30, 2014 at 05:07 PM
"but banks are not obliged to accept any and all deposits. They have no legal obligation to enter into a contract with a prospective customer. However, once they have a legal contract with a depositor, they are obliged to abide by that contract, which includes redeeming deposits for central bank money."
Agreed (apart from illegal reasons such as racial discrimination, etc.). You need to have an existing deposit account or open one up.
I noted that earlier. If you can't get a bank to open up a deposit account for you (e.g. inadequate identification, legal status, etc.), that's a different problem.
Posted by: JKH | March 30, 2014 at 05:45 PM