Speculative musings, from an inadequate historical background.
We tend to take it for granted that the world looks like this, from a monetary perspective:
The world is divided into currency areas. Each currency area has one alpha bank (the king). All the other banks are beta banks (subjects, who follow the king). The beta banks promise to redeem their money for alpha bank money at a fixed exchange rate. The alpha bank makes no such promise the other way. Asymmetric redeemability. The alpha bank is the leader, who decides on monetary policy for the whole currency area. The beta banks just follow along, because they fix their exchange rates to the alpha bank. The world is divided into monetary monarchies.
There may be sub-monarchs, who play alpha to the beta banks within their own sub-area, but who in turn play beta to some greater alpha bank. Like when the Canadian commercial banks fix their exchange rates to the Bank of Canada dollar, but the Bank of Canada in turn fixed its exchange rate to the US dollar. The Canadian commercial banks follow the Bank of Canada, but the Bank of Canada used to follow the US Fed. The Bank of Canada used to be a sub-monarch, before it floated the exchange rate and declared monetary independence, and then became a full monarch.
Does the world have to look like that? Is monetary monarchy inevitable? Is oligarchy an unstable system? Can two or more banks play both alpha and beta to each other, where each promises to redeem its money for the other's money at a fixed exchange rate, and so share monetary sovereignty? Or will monetary monarchy inevitably reappear, where the strongest bank becomes the new alpha bank?
In the couple of years immediately preceding the Euro, and the creation of the ECB, were the central banks of France and Germany true oligarchs, who shared the pooled monetary sovereignty over France and Germany by fixing their exchange rates symmetrically to each other? Could this have continued? Or would one of the pair have become the alpha, and the other the beta, if the ECB had not been created?
Suppose the Bank of Canada disappears. Because people stop using Bank of Canada currency, so nobody ever asks to redeem their Bank of Montreal deposit money for Bank of Canada paper money. And because the Canadian commercial banks stop settling their own accounts on the books of the Bank of Canada, so none ever holds either positive or negative balances at the Bank of Canada. The King is dead. How long before one of the Canadian commercial banks becomes the new King? What determines which of the commercial banks becomes the new alpha central bank? Will it be the strongest bank with the tightest monetary policy? I think it will. The tightest bank will refuse to redeem other banks' money for its money, if that would force it to loosen. The other banks must either follow the tightest, who becomes alpha to their beta, or else declare independence and float their exchange rates. Most banks aren't strong enough to declare independence; nobody will use their money if they do.
Gold used to be the alpha money, and gold miners the nearest thing to an alpha bank. But gold became less used as money, and people used beta paper money that followed alpha gold. Then the beta banks stopped fixing their exchange rates to gold. They stopped following the king, so the old king was no longer king. The strongest of the beta banks, like the Bank of England, became the new kings, within their own currency areas.
Bitcoin is not a beta money, because it does not fix its exchange rate to anything. Is it an alpha money? Does it have beta followers, who fix their exchange rates to Bitcoin?
(Synmetallism, where a bank pegs its exchange rate to a basket of currencies, does violate this simple monarchist perspective. The beta bank follows an average of two or more alphas. And inflation targeting is like synmettalism. The King agrees to be bound by his own laws. (Bimetallism is where you can redeem your money for fixed amounts of gold or silver. Synmettalism is where you can redeem your money for fixed amounts of gold plus silver.))
Worthwhile Canadian Initiative 
I’ve been following the one alpha one beta discussion, and here is how I’m modeling it in my head:
1. Everyone has an electronic account at the beta bank, and uses a debit card for all purchases.
2. The alpha bank issues paper currency, “alpha dollars”, but it only trades that currency to the beta bank. The beta bank is not allowed to trade alpha dollars to the public – it can only trade them bank to the alpha bank.
3. The alpha bank may choose to pay interest on the alpha dollars if it wants.
4. The alpha bank uses the alpha dollars to purchase gold from the beta bank. The alpha bank is not allowed to set both the price and quantity purchased – all trades between the alpha and beta bank are voluntary.
5. The alpha bank has no source of revenue or assets other than purchasing gold from the beta bank using alpha dollars.
6. There is no government; taxes and spending are 0 (or equivalently, taxes and spending are done using deposit accounts at the beta bank).
I think it’s clear in this system that the alpha bank will just cease to exist. There is no reason whatsoever for the beta bank to actually trade with the alpha bank – the alpha dollars it receives can only be used to repurchase the gold traded away. Even if the alpha bank promises to pay interest on the alpha dollars, or (equivalently) reduce the alpha dollar price of gold in the future, that achieves nothing. The beta bank knows that there is no possibly way for it to do any better than if it had just held onto the gold.
If we introduce multiple beta banks, and relax (2) such that the beta banks are allowed to trade alpha dollars among themselves (to e.g. settle accounts), then the alpha bank suddenly has a reason to exist. Banks will trade gold for alpha dollars because those alpha dollars provide positive value.
Note that this continues to work even if we fix the rate of interest on alpha dollars to zero and permanently fix the alpha dollar price of gold (e.g. 1 alpha dollar = 1oz). The alpha bank is basically just a warehouse issuing convenient certificates of ownership for individual gold bars, which are heavy. However, if this alpha bank is required to redeem the alpha dollars on demand then it has no power over monetary policy*.
* I suppose it could have some power in one direction by refusing to issue new alpha dollars
The circle is complete, and we have a real alpha bank that can fully control monetary policy, if we allow it to refuse redemption of alpha dollars.
So I think this does confirm that asymmetric redeemability is required for central banking, but technically not sufficient – asymmetric redeemability doesn’t help in the one alpha / one beta case. Alternatively, we would say the one alpha / one beta case doesn’t have asymmetric redeemability because it has no redeemability at all – and so asymmetric redeemability is something that has to be earned through market power.
JKH,
I don’t think the disappearance of currency makes any difference, providing there are still some other claims on the central bank for which there is sufficient demand. For example mandatory reserve requirements.
So, we have a central bank all of whose liabilities are held by commercial banks. In that case, the redeemability test is whether Bank A can require the central bank to deliver claims on Bank B (or any other commercial bank) in redemption of reserves.
As to whether this is a useful way of looking at things, I would be inclined to agree that it probably wouldn’t figure in a normal description of how central banks conduct monetary policy. I got to it this way. I was thinking about a theoretical situation with several competing private banks, but no currency or central bank. Depositors need to be able to make payments to depositors with other banks, so the banks have to be prepared to take exposure on each other. You need a reasonable interbank market and you end up with a web of interbank exposures. Some banks may in fact not take deposits from the public at all and may fund their loans with capital instruments and interbank borrowing. Now the question is what is needed to give one of those banks the power to dictate interest rates. I came up with the two conditions I listed, the first of which is essentially Nick’s asymmetric redeemability.
Tom,
I agree that if there is a provision that the bank to be wound up if its paper net worth goes negative, then it will need to watch its paper profit. It may also need paper profits to pay dividends.
Alex Godofsky, thanks for your analysis. That’s pretty detailed and clear. I will stew on that a bit. Two questions about the 1-alpha 1-beta scenario:
1. What if the beta bank is taxed in alpha dollars: does that change anything?
2. What if we relax the “voluntary” part of your description, and force (in exchange for alpha dollars + fee) the beta (as part of its monopoly charter) to either:
a) Act as an agent of the alpha to purchase gold from 3rd parties
b) Accept alpha checks for deposit
Would either work? Would either destroy the asymmetry or the alpha nature of the alpha? Why or why not? Thanks.
Also, one other related but different question for you: Same set up as above, but with multiple betas. Say aggregated deposits at the betas amounted to $D, total reserve liabilities of the CB amount to $R, and we’re in equilibrium with a steady state average price level of $P. Also assume D is not equal to R and that R, D and P are all positive. Now the CB purchases all the betas for a steal (say their price is negligible) and takes over banking operations, running the betas exactly like they had been run with no changes. Non-bank entities formerly held beta bank deposits, but now they have alpha bank deposits with the same terms and conditions as before. Does the steady state average price level P change? Why or why not? A simple analysis might say that the total Medium of Account (MoA) changed from R to D, and thus P should eventually change to a new value, P1 = P*D/R. What do you say?
Nick Edmonds, thanks.
O/T: Glasner uses Philippe’s tax argument (used in this post actually) to explain the value of fiat currency:
“…it is possible to account for the positive value of fiat currency by reference to its acceptability in discharging tax liabilities to the government.”
Also it sounds like he thinks Bitcoins are a “bubble.” I’ll be interested to hear Sumner’s rebuttal.
Nick E.,
“So, we have a central bank all of whose liabilities are held by commercial banks. In that case, the redeemability test is whether Bank A can require the central bank to deliver claims on Bank B (or any other commercial bank) in redemption of reserves.”
I´m not following this.
I don´t see how a bank is long the option to redeem reserve balances.
Not long in the way a bank customer is long an option to redeem currency.
What sort of claims on other banks is the central bank delivering?
i.e. it looks like its long the option to use reserve balances as a medium of exchange
which is different than a redemption option I think
JKH,
Say Bank A has a positive balance with Bank C that it doesn’t want to roll; instead it wants to transfer that balance to Bank B. How it does so is not important, but say it does this by simply asking Bank C to pay into Bank B for its (Bank A’s) account. Bank B now needs to cover that payment, so it will have to do something like borrow elsewhere or sell assets.
This is equivalent to Bank C delivering a claim on Bank B in redemption of Bank A’s claim on Bank C. Rather than make a payment, assume Bank C borrows, uses the proceeds to deposit with Bank B and then transfers title to that deposit to Bank A.
Because of its commitment to deliver claims on other banks in redemption of its deposits (whether those of the public or those of other banks), Bank C must pay an interest rate sufficient to prevent redemptions, or to face a reduction in its deposit base.
The point is that there is no equivalent constraint on the central bank, precisely because commercial banks are not long the option to have reserves redeemed in the same way that they are long the option to have claims on other commercial banks redeemed.
Nick E.,
Thanks for your patience in responding to a few questions.
The rest of this comment isn’t directed at you. It’s just a venting on my part.
I think I know a fair about how central bank reserve and clearing systems work – and where central banks get their leverage – and this asymmetry paradigm strikes me as an entirely unnecessary way of getting at the subject.
There are far more straightforward and positive ways to describe how central banks exert their monetary power through quantity and price controls on the monetary base.
The asymmetry meme is a negative specification – identifying something by what it is not. Like defining white as not black. I really question just how constructive this approach is.
The word “redeemability” is itself fraught with ambiguity in use. Banks are said to redeem the deposits of their customers and customers are said to redeem those same deposits. This is perfectly commmon language in banking. Redeemability is two way at the bank-customer interface. There is additional redeemability at the bank-central bank interface and at the bank to bank interface. Optionality is rampant and subtle. This is linguistic chaos unless the accounting entries that correspond to the operational steps are precisely recorded.
More power to Nick R. in looking at it this way. I think you’re more on the sidelines interpreting this particular paradigm, but you are obviously comfortable with the language in doing so.
The paradigm may be ultimately correct in the description, but I very much doubt it represents the core understanding of how central banks exert their monetary power and control. This status is not defined as a negative in my view.
Similar in working up the pipeline to FX and gold pegs.
Perhaps I’ll see it differently some day. But I don’t think so.
I will do a post on the subject at some point.
Thanks again.
Nick Edmonds: “Any bank can buy anything just by creating the deposits.”
Well, the Fed is a funny kind of bank. Congress has limited what it can buy, for instance.
Nick E and JKH,
Let me start with you Nick E, is this a typo?:
“Bank B now needs to cover that payment, so it will have to do something like borrow elsewhere or sell assets.”
Should that be “Bank C?”
Also, you both start using language I’m not 100% familiar/comfortable with (e.g. “long the option”). Not that I’m complaining, but I’d love to see one/both of you do a post on this explaining this in more detail (I got a little lost in both your comments). 😀
JKH, you write “Like defining white as not black” … that wouldn’t work, but you could say white is not a deficiency of any color (whereas black is a 100% deficiency of all colors). 😀
Thanks to both of you for a great interchange!
Also, Min, you write “Congress has limited what it can buy, for instance.”: Mark A. Sadowski once made a list of the dollar amounts and types of assets the Fed could legally buy in a comment. Unfortunately I didn’t save the link: I’m usually pretty good at digging those up, but I don’t know where to start with that one. But maybe he remembers:
http://www.themoneyillusion.com/?p=26587&cpage=1#comment-331974
Or maybe you know a better way to find that.
Min, here you are:
http://www.themoneyillusion.com/?p=26304#comment-322387
From the Woodford paper cited above:
“Why should the central bank play any special role in determining which of these outcomes should actually occur, if it does not possess any monopoly power as the unique supplier of some crucial service?
The answer is that the unit of account in a purely fiat system is defined in terms of the liabilities of the central bank. A financial contract that promises to deliver a certain number of U.S. dollars at a specified future date is promising payment in terms of Federal Reserve notes or clearing balances at the Fed (which are treated as freely convertible into one another by Fed). Even in the technological utopia imagined by the enthusiasts of “electronic money” — where financial market participants are willing to accept as final settlement transfers made over electronic networks in which the central bank is not involved — if debts are contracted in units of a national currency, then clearing balances at the central bank will still define the thing to which these other claims are accepted as equivalent.
This explains why the nominal interest yield on clearing balances at the central bank can determine overnight rates in the market as a whole. The central bank can obviously define the nominal yield on overnight deposits in its clearing accounts as it chooses; it is simply promising to increase the nominal amount credited to a given account, after all it can also determine this independently of its determination of the quantity of such balances that it supplies. Commercial banks may exchange claims to such deposits among themselves on whatever terms they like. But the market value of a dollar deposit in such an account cannot be anything other than a dollar — because this defines the meaning of a “dollar”! This places the Fed in a different situation than any other issuer of dollar-denominated liabilities.”
Click to access w8674.pdf
Phillipe’s post said: “”Assume a 5% risk-free overnight rate” On what, Reserves? Why?
“The alpha bank buys a bond/sells central bank reserves”
why does the commercial bank accept the central bank reserves? I don’t get the point of the rest of your comment, sorry :)”
The risk-free overnight rate should affect rates on short-term treasuries, rates on savings accounts, and rates on CD’s.
Why does the commercial bank accept the central bank reserves?
1) The central bank overpays for the assets.
2) That is the way the system is set up.
Same question could apply to QE?
JKH, can the monetary base be zero, and the central bank still control the risk-free overnight rate (fed funds rate)?
I don’t see why not.
Tom,
Yes, that’s a typo. Thanks. I’m going to have to start asking you to proof read all my comments before I post them!
I’m not sure the option terminology needs a post. Basically an option here is a contract between two parties to exchange one asset for another asset at a fixed exchange rate, where only one party gets to choose whether the exchange takes place or not. The party that gets to choose is long the option; the other party is short.
JKH,
No worries. I will look forward to your post on this.
TMF,
Required reserves can be zero.
And currency can be zero in theory.
But actual reserve balances must be positive at some point during the day.
This is because intraday payment clearing in batches virtually ensures that some bank will be short balances on an intraday basis.
So the central bank lends intraday by way of daylight overdraft as necessary.
The intraday central bank balance sheet will show loans of reserves and reserves at a minimum.
In theory the overnight monetary base could clear to zero.
This is Woodford in the paper cited above on the special role:
“Why should the central bank play any special role in determining which of these outcomes should actually occur, if it does not possess any monopoly power as the unique supplier of some crucial service?
The answer is that the unit of account in a purely fiat system is defined in terms of the liabilities of the central bank. A financial contract that promises to deliver a certain number of U.S. dollars at a specified future date is promising payment in terms of Federal Reserve notes or clearing balances at the Fed (which are treated as freely convertible into one another by Fed). Even in the technological utopia imagined by the enthusiasts of “electronic money” — where financial market participants are willing to accept as final settlement transfers made over electronic networks in which the central bank is not involved — if debts are contracted in units of a national currency, then clearing balances at the central bank will still define the thing to which these other claims are accepted as equivalent.
This explains why the nominal interest yield on clearing balances at the central bank can determine overnight rates in the market as a whole. The central bank can obviously define the nominal yield on overnight deposits in its clearing accounts as it chooses; it is simply promising to increase the nominal amount credited to a given account, after all. It can also determine this independently of its determination of the quantity of such balances that it supplies. Commercial banks may exchange claims to such deposits among themselves on whatever terms they like. But the market value of a dollar deposit in such an account cannot be anything other than a dollar — because this defines the meaning of a “dollar”! This places the Fed in a different situation than any other issuer of dollar-denominated liabilities.
This is a test of the spam filter with Ghostery and Firefox.
Test of the spam filter with Chrome.
My comments increment the comment count and appear in the recent comment list, but they don’t appear in the actual comments. Very strange.
Peter N, before me writing this comment, you had the last four comments in the recent comments list.
Nick Rowe, O/T: I thought of you when I saw this, with your repeated analogy that to get a broom balanced upright in the palm of your hand to move one direction you have to start off moving your hand the opposite direction:
Probably even better illustrated in the first minute or so here (for the case of three brooms balanced one on top of one another):
The best part is those guys have it all written down mathematically as a tidy feedback/feedforward control law… now just to translate to macro. 😀
This one’s pretty amazing too, especially once you realize how crazy an uncontrolled double pendulum can be:
first uncontrolled:
controlled:
This one’s pretty amazing too, especially once you realize how crazy an uncontrolled double pendulum can be:
first uncontrolled:
controlled:
This one’s pretty amazing too, especially once you realize how crazy an uncontrolled double pendulum can be:
first uncontrolled:
controlled:
JKH said: “Required reserves can be zero.”
OK
“And currency can be zero in theory.”
OK
“But actual reserve balances must be positive at some point during the day.
This is because intraday payment clearing in batches virtually ensures that some bank will be short balances on an intraday basis.”
What if there is only one commercial bank?
JKH said: “Required reserves can be zero.”
OK
“And currency can be zero in theory.”
OK
“But actual reserve balances must be positive at some point during the day.
This is because intraday payment clearing in batches virtually ensures that some bank will be short balances on an intraday basis.”
What if there is only one commercial bank?
JKH said: “Required reserves can be zero.”
OK
“And currency can be zero in theory.”
OK
“But actual reserve balances must be positive at some point during the day.
This is because intraday payment clearing in batches virtually ensures that some bank will be short balances on an intraday basis.”
What if there is only one commercial bank?
TMF, it looks like you encountered the same error message when you tried to post: I think the trick is to just reload the page, and then your comments appear
TB, I’ll try that.
“Suppose the Bank of Canada disappears. Because people stop using Bank of Canada currency, so nobody ever asks to redeem their Bank of Montreal deposit money for Bank of Canada paper money. And because the Canadian commercial banks stop settling their own accounts on the books of the Bank of Canada, so none ever holds either positive or negative balances at the Bank of Canada. The King is dead.”
Suppose the Bank of Canada remains. People stop using Bank of Canada currency, so nobody CURRENTLY ever asks to redeem their Bank of Montreal deposit money for Bank of Canada paper money. They could change their mind. Bank of Montreal is the only commercial bank, and there is a 0% reserve requirement. The Bank of Montreal has no vault cash and no central bank reserves. That is both the desired and required amount. The Bank of Canada has no assets and no liabilities. The Bank of Canada can set the risk-free overnight rate with a bond purchase or bond (short) sale, which it will reverse when the risk-free overnight rate is where it wants it to be. The Bank of Canada will allow new currency creation if the assets perform or are expected to perform. The Bank of Canada also sets the minimum capital requirement.
Now run the model.
“Donald,
I guess then that banks were not offering par convertibility between bank notes either then. Otherwise, if you couldn’t repay Bank A with Bank B’s notes, you’d just go to Bank A (or Bank B) and swap your Bank B notes for Bank A notes, then repay your loan. Likewise, if everyone’s offering fixed rate convertibility to gold, you convert your Bank B notes to gold with Bank B and then the gold for Bank A notes with Bank A, then repay your loan.”
You could do that, but nobody ever offered par convertibility between different cities in the 19th century US.
There are sections in 19th century math textbooks regarding conversion of New York banknotes to New Orleans banknotes — there was always a small discount, a conversion fee. You had to keep track of how much your money was worth at different locations.
You could convert to and from gold at par, but then you had to haul the gold from New York to New Orleans, which ended up costing more.
Banks within the same city usually ended up offering par convertibility, precisely because it was practical to withdraw gold from one bank and walk across the street to deposit it to the other bank.
Interestingly, there are cryptocurrencies which are explicitly trying to ‘out-alpha’ bitcoin by irreversible, permanent destruction of bitcoin called ‘proof of burn’. Hasn’t caught on too far but enough to be noteworthy in this context, as it’s an explicit damage to the bitcoin stock(potentially causing deflationary effects, on an already relatively deflationary currency, even)
“Does it have beta followers, who fix their exchange rates to Bitcoin?”
Most of the cryptocurrencies available on cryptsy and other platforms seem to have no value other than as methods of either
a) storing bitcoin wealth off blockchain (up to and including as a method of ‘laundering’ coins)
b) have one particular use case, which is not a strong enough case for independence from bitcoin(for example, freicoin). These coins have their value defined almost entirely by their relation on open markets to the bitcoin.
for a list of cryptocurrencies, see https://coinmarketcap.com/ of which there are about 283, most of which are functionally ‘betas’ to bitcoin, by my reckoning at least.