« Why are groceries more expensive in the bulk food section? | Main | Raising the Inflation target vs Depreciating Paper money against Silicon money »

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

Is the trick supposed to be that this is already the case (that alpha silicon "appreciates" relative to alpha paper) because alpha paper has annoying carrying costs?

(Alternately, if / because the alpha bank pays interest on alpha silicon?)

In the case where this is literally just an implementation in terms of the exchange rate (and not due to carrying costs or interest) then I'd expect the peg with alpha paper to stick and for real economic transactions to be priced in terms of alpha paper. I'd expect consumers to avoid using alpha silicon for much, and for it to be relegated to settlements between the beta banks, who can effortlessly do the multiplication when they need to.

I'd expect the above because alpha paper's main use is retail and multiplying would be very annoying (and there are no beta substitutes), whereas it is easy for people to do all of their own banking using beta silicon.

Alex: There aren't supposed to be any tricks.

Alpha paper does have small storage costs, so you can have slightly negative interest rates on alpha silicon before alpha paper dominates. But if the alpha bank announces a crawling peg, in which paper depreciates against silicon, it can set much bigger negative rates on silicon.

Ah. In that case then yes, I expect the crawling peg to be unhelpful because having to set one price for beta silicon and one price in alpha paper would be really annoying to pretty much everyone - workers, consumers, businesses.

This means that any nominal stickiness or money illusion will be in units of the alpha paper and so you don't get anything out of negative rates on alpha silicon, unfortunately.

Humble I/O guy joins:
Nick :"Alpha paper is real physical stuff, that even kids understand. That might make it salient."
How long? My father, who founded a Caisse Populaire (a kind of credit union for those outside franco QC) once told me that,in the 50's, when warned that their checking account was overdrawn, people wrote a check and deposited it. Even the most uncultured would not do that today. My little nephews seem to have a rather good grasp of what's behind a debit card. (The 5 year-old once told me that "a credit card is to pay for things and get into debt." And he knew what a debt was.
Implications for LT stability of monetary-banking policy? Is the cat condemned to chase its tail?

Alex: you might be right on that. Unless Jacques Rene is right, and the kids really do think differently from us old folks, who apologise profusely before paying silicon for a $1.50 double double when we find our pockets are empty. (My Grandmother never got her head around metrification of Sterling, and would only talk about pounds, not pence, thereafter.)

Oh, it's definitely the case that alpha paper is very marginal as a medium of exchange. But it's used a lot more often than alpha silicon (outside of the sort of interbank transactions where automatically applying an exchange rate conversion is cheap and easy).

Having to actually multiply the price by some factor to figure out how many slips of alpha paper to pull out would be incredibly annoying and would downgrade alpha paper from "marginal" to "almost useless".

Alex: very good point, which I should have made myself. Have updated post.

"If alpha paper is salient, then Miles Kimball's plan to get sharply negative rates on silicon to create inflation won't work. From a paper standard, if the alpha bank charges 1% negative interest on alpha silicon, and at the same time announces a crawling peg in which alpha silicon appreciates at 1% against alpha paper, it's a wash. Nobody cares."

Yes, exactly. Getting everyone onto an electronic money standard is one of the challenges that Miles's plan faces.

He usually points out that the central bank would have to intervene to get everyone on the same electronic standard, in the same way that the government coordinates everyone for daylight savings, or how the ECB staged the introduction of the euro unit of account.

Another point is that we may already be on a silicon, or electronic dollar standard. It's not always obvious, but retail prices are set as-if a credit card (ie deposit) is being used while cash payments get a rebate.

If we had a paper money oriented economy like Zimbabwe, then yeah, it would be natural for the paper dollar to be the unit of account, no matter what.

As it is, I think you could do it, but not without legal changes. At a minimum, you have to revoke legal tender status for paper money.

"[It's the stupid framing of monetary policy as nominal interest rate policy that is to blame for leading us into philosophical problems like this.]"

What about the confusion between central bank money and commercial bank money?

The central bank issues money - it has (theoretically) a liability.

A commercial bank issues money (i.e. a loan) - it has an asset.

I think there's one factor missing in this world: why would beta banks ever hold alpha silicon?

Ordinary people can't use alpha silicon, so it's not walking through the door. You haven't posited a way by which the alpha bank forces beta banks to hold its silicon. The only endogenous use is for transaction settlements in lieu of briefcases full of alpha paper, but even here many transactions will net out such that beta banks only need alpha silicon for imbalanced flows.

Outside of this small transactional demand, banks will only want to hold silicon if there is profit in doing so. That profit derives from their privileged positions of being the only ones able to hold alpha silicon.

If alpha silicon is appreciating against paper, then banks will want to attract paper such that beta silicon circulates and the banks make profit on the spread. Beta silicon may or may not also appreciate against paper, and the MoA is still alpha paper because there is no standardized beta silicon appreciation rate (differing between banks).

If alpha silicon is depreciating against paper, then banks will want to… I'm not sure. Hold negative alpha silicon? Are we letting banks do that?

The flip side of this argument is that there's also no private credit market, which would seem to be crucial to the actual economy. Beta silicon here is only used as a payment intermediary, with nobody holding negative positions for very long.

It would vary with the economy. One predominately paper based would stay with paper while a modern one where 2/3rds of transactions are already electronic would use an electronic standard and further push paper out of the way and cash registers would handle the conversion. Paper users would just have to expect they will need more.

Alex,

As I understand it the alpha silicon is the equivalent of reserve balances of banks. If that's true then I am not sure how your statement can be true:
"Oh, it's definitely the case that alpha paper is very marginal as a medium of exchange. But it's used a lot more often than alpha silicon (outside of the sort of interbank transactions where automatically applying an exchange rate conversion is cheap and easy)."

The Fed processes payments of $82 billion per day or $20.5 trillion per year through their reserve system: https://www.federalreserve.gov/paymentsystems/fedach_yearlycomm.htm
I doubt you get that volume with cash.

One initial premise is that the alpha bank issues silicon usable only by the beta banks. If alpha silicon is usable only by beta banks, how would beta banks find any value in it?

Until we know the reason that beta banks find value in alpha silicon, we seem to have an unsolvable question.

On the other hand, if alpha bank was willing to trade beta silicon for alpha silicon, then beta banks could use alpha silicon as a storage mechanism. This would only have value if there was a cost to store beta silicon in beta owned silicon storage.

Odie:

"outside the sort of interbank transactions..." is the key phrase there. All of that volume is, as I understand it, stuff like settlement between banks or other transactions between sophisticated financial institutions. And, of course, computerized. So it would be virtually costless to have to apply some multiplicative factor when performing those since they're automated anyway.

For the first part of this--up to the point at which Alpha Bank starts to depreciate its paper money--there might be much to be learned from the US experience in the years before the establishment of the Fed (and in particular during the period in which there was a quasi-central bank, the first and second Banks of the United States. There was an alpha currency--gold and silver coins--and the beta banks (state-chartered) were supposed to be able to redeem their paper currency at par with gold and silver coins (i.e., $10 of my bank's bank notes were supposed to be redeemable as 10 silver dollars or one $10 gold piece). The first and Second Banks of the United States were the enforcers, accepting beta bank notes and then returning them to those banks for redemption in gold or silver coins. It's a fascinating period in US banking (and monetary) history.

JP: Yes, the Daylight Saving Time analogy shows it *can* be done. (Then there's the time they changed the calendar, and some people insisted the government/church was shortening their lives by X days??)

Max: But it would be tricky to make alpha silicon "legal tender" (whatever that means in various countries) if nobody uses it except banks. Hmmm.

Majro: Canadian banks normally hold very little alpha silicon. They can hold positive or negative balances (green or red alpha silicon), but since the interest rate on positive is always 50bp below that on negative, they normally lend to each other overnight at around the midpoint 25bp between the two rates. But it doesn't really matter if green alpha silicon = red alpha silicon summing over banks.

Lord: yep. good point. It would be very hard to switch to a silicon standard if most people actually used paper.

Roger: beta banks value alpha silicon because it is a medium of exchange for them. Plus it's convertible into paper.

Donald: Yep, looking at history might tell us what to expect. I am also reminded of Harold Wilson's "Pound in your pocket" speech, when the UK devalued.

Nick:

You say "Roger: beta banks value alpha silicon because it is a medium of exchange for them. Plus it's convertible into paper."

OK, but what do beta banks trade to the alpha banks to acquire alpha silicon? Beta silicon?

Alex,

Sure, those are interbank settlements but most of them happen on behalf of the customers of those banks. When customer A wants to transfer a certain sum to customer B at another bank the banks will exchange the same amount in reserves to facilitate the exchange. Those "interbank settlements" are the bread-and-butter of our economy and used for automated deposit transfers or check clearing.

And with your second sentence do you mean there is some "exchange factor" applied when doing the transfer? I am scratching my head how that is going to work. Someone transfers their $1000 rent to their landlord and he/she only receives $990?

"Question 1: will the beta banks maintain a fixed exchange rate between beta silicon and alpha silicon, or between beta silicon and alpha paper?"

But don't they all have different exchange rates in the sense that they all have different interest rates (meaning crawling peg) attached (eg. IOR != interest on check accounts != 0)? I'm not sure anything is profoundly shaken if interest on paper is not equal to zero. But certainly demand for silicon and paper will be affected.

Question 2: will the unit of account be the silicon dollar or the paper dollar?

I think it is a combination of these. Neither is pure medium of account but their "cheapest-to-deliver" is. Also there might be different charges (or discounts) at the counter in addition to interest rates attached.

If beta banks create silicon money why do they want alpha silicon?

In the current world these silicon monies are the same.

So when does anyone ask for alpha silicon? Even if the alpha bank says the alpha silicon must be purchased by more alpha paper today than yesterday, a person/non-bank is not able to buy alpha silicon, something missing in the story about why beta banks care about alpha stuff (especially if they already own huge quantities of alpha silicon).

JF,

Besides interbank transfers the treasury uses alpha silicon to receive its tax payments and to spend (at least in the US and Germany; I assume also in Canada). To say it more applicable to the real world: Governments usually hold their accounts at the central bank. Not using alpha silicon therefore means for a bank that it could not process any transfer that involves the treasury.

Suppose a CB maintain 2 currencies and it wants to target different own-rates of inflation on each. It is able to offer an own-rate-of-interest for holding one currency (this can be either positive or negative) while on the other the own-rate is always nominal zero.

Just by adjusting the 2 quantities of currency on a trial and error basis (and by letting the rate of interest on the first currency float to the market-clearing rate) wouldn't it eventually hit the right volumes of each currency so that each would hit their own-rate-of-inflation target ?

Alternatively it could say it has 2 instruments
1) The rate of interest on one currency
2) The rate of exchange it offers between the 2 currencies.

By adjusting these 2 variables I think it could also always hit the 2 targets.

And if can always hit the 2 targets doesn't that mean it can always control inflation no matter how it chooses to measure it, and no matter which currency ends up being used as the unit of account, or used by beta-banks as base-money for their loans?

I do not agree with this:

"[It's the stupid framing of monetary policy as nominal interest rate policy that is to blame for leading us into philosophical problems like this.]"

because of this:

"Note that the alpha bank does not promise to convert beta silicon into alpha paper or silicon, which is what creates the asymmetry between alpha and betas."

As long as the 5 beta banks are solvent, the alpha bank will act (if necessary) as an unlimited lender of last resort to the 5 beta banks even if it means violating the price inflation target.

Elastic (paper) currency is given priority over the price inflation target.

Odie said: "Besides interbank transfers the treasury uses alpha silicon to receive its tax payments and to spend (at least in the US and Germany; I assume also in Canada)."

Is that right?

Let's assume I have a checking account at bank A. The gov't has a TT&L account at bank A. Firm B has a checking account at bank A.

The gov't taxes me $1,000. I pay it. The gov't spends the same $1,000 at firm B. The beta silica "moves" from me to the TT&L account and then "moves" to the checking account of firm B. No alpha silica involved.

TMF,

In principle you are right. TT&L accounts were instituted to allow banks manage their overall reserve positions better and not having the treasury being a steady drain on those reserves. Nevertheless, a bank must also be able to deliver the funds to the treasuries reserve account: "Under one such program, collector depositories transfer some payments to a Treasury account at the Federal Reserve the same day they receive them, and transfer all other payments to the Treasury's account a day later." https://www.newyorkfed.org/aboutthefed/fedpoint/fed21.html Hence, a reserve account is a prerequisite for a bank to deal with the treasury. If a bank is not part of the reserve system it needs to have a correspondence bank that is and can clear such payments for them.

And deposits do not just move from bank A to bank B. Bank A has to transfer an asset to bank B in order for it to credit that firm B with a deposit (which is a liability of bank B). That asset is usually moving a corresponding reserve deposit from bank A to bank B. It is easier when you draw up the balance sheets of the banks involved. Let's say customer A directly pays firm B at bank B $1000. Bank A debits $1000 from the customer's account (extinguishes its liability). Then it transfers $1000 from its reserve account to the reserve account of bank B (asset transfer). Bank B will now credit the account of firm B with the $1000 (create new liability). Bank A's balance sheet shrunk by $1000, bank B's balance sheet increased by $1000. However, an equal amount of assets and liabilities were "exchanged" so the capital positions of both banks stays the same meaning none made a profit or loss directly on that transfer (assuming no fees etc). Although, bank B may now be able to get interest for that reserve balance it received which would increase its assets and therefore its capital position later on.

That mechanism is what I suggested in my post to Alex. $20 trillion per year cleared in the interbank market through the reserve system happen because of customers moving deposits from one bank to another.

Odie:

Sure, those are interbank settlements but most of them happen on behalf of the customers of those banks. When customer A wants to transfer a certain sum to customer B at another bank the banks will exchange the same amount in reserves to facilitate the exchange. Those "interbank settlements" are the bread-and-butter of our economy and used for automated deposit transfers or check clearing.

And with your second sentence do you mean there is some "exchange factor" applied when doing the transfer? I am scratching my head how that is going to work. Someone transfers their $1000 rent to their landlord and he/she only receives $990?

Right but let's say I send you $1000 in beta silicon electronically and we use different banks. And let's say the exchange rate is that this is $900 in alpha silicon. My bank's website is going to prompt me for how much I want to send you, I type "1000", and then it sends over $900 alpha silicon to your bank, which credits your account with $1000 beta silicon. The translation doesn't have to be visible to me at all and can be just part of the computer process, no more burdensome than programming the two computers to talk to each other in the first place.

Alex,

Sorry, my question was imprecise. It would certainly be technically possible to apply a multiplier to any reserve transfer. I am wondering what that will mean for the banks and their customers. In your example, bank B would only receive an asset of $900 but create a liability of $1000 in my account. That means the bank's capital will be reduced by $100. Essentially, the bank's owner(s)/shareholders will "donate" those $100 to me. I find that behavior rather unlikely for a profit-maximizing business so what will probably happen is an increase in the amount you will need to transfer from your account to cover the multiplier. ($1111 debit from your account assuming a 10% exchange multiplier.) In that case, bank B would receive a $1000 reserve deposit which will result in a $1000 credit to my account.

Negative interest rates on reserve balances are ultimately the same as a tax on bank capital. If applied during reserve transactions with a negative "multiplier" and then passed on to customers you have the equivalent of a sales tax on deposit transfers. The Fed seems to agree: https://www.stlouisfed.org/on-the-economy/2016/may/negative-interest-rates-tax-sheep-clothing

It seems counterintuitive to me that tax increases will stimulate economic activity but considering that negative interest rates on reserve balances are already being used in some countries my intuition is probably wrong.

Odie said: "In principle you are right. TT&L accounts were instituted to allow banks manage their overall reserve positions better and not having the treasury being a steady drain on those reserves."

And to help manage the fed funds rate for the fed?

See:

https://www.newyorkfed.org/medialibrary/media/research/current_issues/ci18-3.pdf

Odie said: "Besides interbank transfers the treasury uses alpha silicon to receive its tax payments and to spend (at least in the US and Germany; I assume also in Canada)."

It looks to me like the treasury taxes beta silica (mostly) and spends with beta silica. The alpha silica is only used to "clear the payments".

"And deposits do not just move from bank A to bank B. Bank A has to transfer an asset to bank B in order for it to credit that firm B with a deposit (which is a liability of bank B). That asset is usually moving a corresponding reserve deposit from bank A to bank B. It is easier when you draw up the balance sheets of the banks involved. Let's say customer A directly pays firm B at bank B $1000. Bank A debits $1000 from the customer's account (extinguishes its liability). Then it transfers $1000 from its reserve account to the reserve account of bank B (asset transfer). Bank B will now credit the account of firm B with the $1000 (create new liability). Bank A's balance sheet shrunk by $1000, bank B's balance sheet increased by $1000. However, an equal amount of assets and liabilities were "exchanged" so the capital positions of both banks stays the same meaning none made a profit or loss directly on that transfer (assuming no fees etc). Although, bank B may now be able to get interest for that reserve balance it received which would increase its assets and therefore its capital position later on.

That mechanism is what I suggested in my post to Alex. $20 trillion per year cleared in the interbank market through the reserve system happen because of customers moving deposits from one bank to another."

I believe the balance sheets end up the same if some entity withdraws alpha paper from bank A and then redeposits it at bank B.

Now I am also going to say the same idea applies between the beta banks themselves so that alpha silica is only used to “clear the payments” and along with the reserve requirement to set the fed funds rate.

With that idea and along with unlimited lender of last resort to solvent beta banks means alpha silica is not MOE. Alpha silica is not MOA either.

Nick:

“Note that the alpha bank does not promise to convert beta silicon into alpha paper or silicon, which is what creates the asymmetry between alpha and betas. The alpha bank is the hub of a hub-and-spoke system of fixed exchange rates, where each beta is responsible for fixing its exchange rate to the alpha. The alpha is responsible only for fixing the exchange rate between its own two media of exchange.”

I’ve always found your premise of asymmetric redeemability difficult and counterintuitive, and something I instinctively disagree with.

For example, BMO issues negotiable paper representing its deposits in effect – by issuing claims on its deposits in the form of cheques. When the Royal Bank presents a BMO cheque to the Bank of Canada in the clearings, the Bank of Canada fixes the exchange rate for converting that BMO cheque (i.e. beta money) into “alpha” reserve balances. That is a promise fulfilled. In the context of redeemability, that is symmetric to BMO fixing the exchange rate for the conversion of beta deposit money into “alpha” paper.

There is definitely a difference in institutional interface for redemption in the two directions - due to the asymmetric clearing structure inherent in central and commercial banking as a system (the asymmetry of hub and spokes in itself).

But there is not really a difference or an asymmetry in the aspect of core redeemability of alpha money and beta money in both directions.


JKH, let me try an example.

Assume an ecb type central bank, one gov’t, and one commercial bank. Also, assume all entities use currency (alpha paper) on even days and demand deposits (beta silica) on odd days and the commercial bank is solvent at all times.

Start in “equilibrium” so that the price inflation target is met. Day 1 every entity spends with demand deposits. Price inflation target met. Day 2 every entity converts demand deposits to currency. Central bank allows total conversion. Entities spend. Price inflation target met. Day 3 every entity converts currency to demand deposits. Entities spend. Price inflation target met. This goes on for awhile.

Day 25 entities borrow from the commercial bank so that demand deposits increase. Entities spend with all demand deposits. Price inflation goes above target.

Day 26 I say entities convert demand deposits to currency. Central bank allows total conversion (unlimited lending of last resort to the solvent commercial bank). Entities spend with all currency. Price inflation in terms of currency also goes above target. Central bank raises fed funds rate.

Day 26, I *think* nick would say total conversion by the central bank does not happen because that would violate the price inflation target. Not sure what happens next. Fed funds rate rises because of shortage? Commercial bank calls in loans? Other? I *think* there would also be two prices (say $1 of currency for an item and say $1.10 of demand deposits for the same item).

JKH: As I was writing that bit, I was wondering if you would chime in, and am glad you have. Because when the BoC issues both paper and silicon, it complicates my original story.

In a world where the BoC issues currency only, it seems clear. It is BMO's responsibility to ensure that BMO paper is convertible into BoC paper.

Suppose BMO refused to honour its commitment to exchange BMO silicon for BoC paper at par. Would it still be allowed to convert BMO silicon into BoC silicon (use the clearing house)?

Should be: "It is BMO's responsibility to ensure that BMO silicon is convertible into BoC paper."

Nick,

I interpreted your original story and I think you state it as the central bank issuing both silicon and paper, with the commercial banks using CB silicon for interbank clearing payments while issuing their own silicon deposits. In reality, they also act as distributors of CB paper, with the usual exchanges of paper for silicon and vice versa along that pipeline using CB silicon and commercial bank silicon in exchange for CB paper.

“If the BoC issues currency only …”

I think the general point I was trying to make still holds. I assume the commercial banks make payments to each other in the clearings using paper money (which is interesting in part because I think that’s a schematic you’ve often used in discussing other aspects of bank reserves). The commercial banks furthermore issue silicon deposits to their customers and exchange silicon deposit liabilities with their customers for CB paper and vice versa on demand.

Suppose the Royal brings a BMO cheque to the Bank of Canada in the clearings. That cheque is in effect a representation or claim on BMO silicon money.

The Bank of Canada will acquire that silicon claim issued by BMO by paying Royal with Bank of Canada paper money (under your clearing payment assumption). The Bank of Canada is in a position relative to Royal analogous to Royal’s position relative to a Royal Bank household customer, for example. In both cases, the customer will be paid in paper money in exchange for cashing out from a silicon bank deposit claim. In the clearing case, the payment of paper money to Royal reflects Royal’s corresponding creation of its own silicon deposit liability when the BMO silicon claim was deposited to that Royal silicon account, followed by the subsequent exchange of the BMO silicon claim for paper money.

The Bank of Canada will then present that silicon claim back to BMO and demand payment in paper money issued by the Bank of Canada. This time, the Bank of Canada is in a position relative to BMO analogous to BMO’s position relative to a BMO customer. In both cases, the customer will be paid in silicon money in exchange for cashing out of a paper money claim. In the clearing case, the payment to BMO of BMO's own issued silicon money claim allows it to extinguish a corresponding silicon deposit liability.

So I see some symmetry in that description - which is why I’ve thought that your take on redeemability is not the best characterization of the alpha/beta construct. (I noticed that you hesitated just a bit on this language yourself in your interview with David Beckworth, but in addition I think there’s a different kind of asymmetry in substance than what you’ve described.)

There’s definitely some interesting asymmetry in central banking; I’m just not sure that asymmetric redeemability is the best characterization of it.

On your second point:

“Suppose BMO refused to honour its commitment to exchange BMO silicon for BoC paper at par. Would it still be allowed to convert BMO silicon into BoC silicon (use the clearing house)?”

I’m not sure about dishonouring a commitment, but I think the second part would be effective in a “cashless” monetary system.

JKH said: "The commercial banks furthermore issue silicon deposits to their customers and exchange silicon deposit liabilities with their customers for CB paper and vice versa on demand."

I don't think that answers the question.

I *think* Nick is saying that the price inflation target applies to exchanging demand deposits for currency. That means the fixed exchange rate applies, but the central bank may not convert all demand deposits to currency. If that is what he means, is that right?

In a world where the BoC issues currency only, it seems clear. It is BMO's responsibility to ensure that BMO paper is convertible into BoC paper.
Suppose BMO refused to honour its commitment to exchange BMO silicon for BoC paper at par. Would it still be allowed to convert BMO silicon into BoC silicon (use the clearing house)?

I think one could imagine a world where BoC silicone was a completely different currency from commercial bank currency, even with a floating exchange rate. Not practical, but imaginable. That would eliminate the responsibility of any bank to convert its silicone into BoC silicone. The question then becomes who is responsible for making sure that bank 1's silicone trades at par with bank 2's etc.? After all, a wheel with one spoke does not go 'round. And seing as in such a hypothetical world there are two exchange rates between money and goods, which would count towards inflation?
I say a CB's responsibility is making sure bank money (credit) trades at par, mainly by setting and enforcing lending standards, while influencing overall credit volume such that bank money retains a desired exchange rate with real goods and services. Paper money can be added afterwards.

Sorry, tags got mangled on my phone. Rough train ride...

Nick

My first comment was on the background point of “asymmetric redeemability”.

This comment goes more directly to the subject of your post.

“If alpha paper is salient, then Miles Kimball's plan to get sharply negative rates on silicon to create inflation won't work. From a paper standard, if the alpha bank charges 1% negative interest on alpha silicon, and at the same time announces a crawling peg in which alpha silicon appreciates at 1% against alpha paper, it's a wash. Nobody cares.”

I disagree with “won’t work”.

Suppose at first the medium of account is silicon.

And the unit of account is $ 100 silicon.

Suppose the central bank sets interest rates on silicon at (1) per cent.

And the result is 2 per cent inflation in the silicon price level.

So the price level increases from $ 100 to $ 102.

Assume the exchange rate between silicon and paper starts out 1:1.

So $ 100 in paper is worth $ 100 in silicon at the beginning of the year and $ 99 in silicon at the end of one year.

So the ending price level translated to paper is:

$ 102 x (100/99) = $ 103 approximately.

And the paper inflation rate is 3 per cent.

Paper holders would have to pay $ 103 in paper for the same goods and services.

Now just change the medium of account to paper. We know already that the central bank has taken action on silicon interest rates to produce an inflation in prices as measured in silicon. It is perfectly reasonable to think that vendors and the banking system will arrange to permit payments in either medium of exchange – just as they do in the real world with respect to cheques or credit cards or debits cards on commercial bank silicon and central bank paper. And there is no reason to assume any different outcome in the case of a change in the medium of account, in terms of the translated value of the price level in either money form according to this central bank action. If at the outset the assumed inflation rate in silicon terms is 2 per cent, there is no reason for vendors to translate a posted silicon price to a posted paper price at a rate that is any different than what the effect of combining silicon inflation with a paper/silicon exchange rate would determine by straightforward math.

So if the medium of account was paper, the posted price would be $ 103 instead of $ 102.

Therefore I think the choice for medium of account makes no difference, assuming liquid markets in the exchange between silicon and paper, and flexible medium of exchange arrangements as is now the case.

(I think other commenters and/or you may have raised issues regarding such assumed flexibility, but this aspect is not directly relevant to the point made in your quote above.)

Finally, it is interesting why the measured nominal inflation rate is different for the two monies.

Paper is analogous to foreign exchange in this case.

And the exchange rate is determined in a way that is analogous to “covered interest arbitrage” in the case of foreign exchange.

The way in which the central bank manages a crawling peg is analogous to the way in which the foreign exchange markets would determine a forward foreign exchange rate – based on an assumed interest rate differential of 1 per cent over the year. That differential then becomes embedded in the determination of the price level expressed in foreign exchange terms.

A comparable embedded money value doesn’t happen in the way in which domestic inflation measures are calculated. There is no adjustment for the nominal value of money, which doesn’t change.

Still - there does exist an analogous domestic inflation calculation.

In this case, you can calculate the opportunity cost of holding silicon money for one year at (1) per cent – which is simply 1 per cent considered in isolation of other factors. When you combine that with the posted inflation rate of 2 per cent, you can arrive at an “interest-rate-adjusted inflation rate” of 3 per cent in silicon terms – the same as the paper case.

I think there is a logical error in your statement that I quoted above. I’m not sure how to describe it other than what I’ve just said. However, it bothers me that JP Koning has agreed with you most heartily.

Odie:

Sorry, my question was imprecise. It would certainly be technically possible to apply a multiplier to any reserve transfer. I am wondering what that will mean for the banks and their customers. In your example, bank B would only receive an asset of $900 but create a liability of $1000 in my account. That means the bank's capital will be reduced by $100.

No the bank receives an asset of 900 alpha silicon which is worth 1000 of any of the beta silicons.

It's the same as if instead of alpha silicon it was gold coins, and 1 beta silicon was worth 0.9 gold coins, and gold coins could be transmitted electronically.

JKH said: "So if the medium of account was paper, the posted price would be $ 103 instead of $ 102."

Does that mean the posted price would be $103 in terms of alpha paper and $102 in terms of beta silica?

JKH: I think I see your point. It makes me realise I should have made a different argument. It's not about raising the inflation rate; it's about avoiding the ZLB while keeping the same (say 2%) inflation target.

Suppose the alpha bank refuses to raise the inflation target above 2%, because it believes that "menu costs" (the costs of changing prices) would be "too high" if it did so. If silicon becomes the unit of account, then Miles plan works. But if paper becomes the unit of account, then the alpha bank faces a dilemma: a 2% silicon inflation target means 3% paper inflation, and menu costs are too high; but a 2% paper inflation target means the ZLB is just as binding as before.

(Though there is a quite different point, about coordination games and inflation inertia, and trying to break that expectational inertia and raising the inflation rate, that I was making originally. It's hard to get everyone to switch to driving on the other side of the road, unless everyone believes that everyone else will switch at the same time.)

But my brain isn't working very well at the moment. And sometimes I like to digest thoughts slowly, and let them reappear a few months later.

Nick,

It seems intuitive that a silicon medium of account would be more convenient in the circumstances, and perhaps there are more refined economic arguments supporting that. My point was merely that I see no reason why either form of money in theory couldn’t serve as the medium of account. It would just require a liquid exchange market between silicon and paper, and a translation facility on the part of vendors. This all seems reasonable when comparing it with the starting point where in today’s world silicon and paper forms are both readily accepted (obviously absent paper depreciation, so far).

I perused some of Kimball’s writing again very quickly, and while seeing that he assumes an electronic medium of account (as JP Koning stated), I didn’t see an argument as to why that needs to be the case in his view.

The behaviour of the two inflation rates together is interesting. If the central bank eases in the form of negative rates and a managed depreciation, the paper inflation rate moves a step ahead of silicon inflation. So it exaggerates the inflation measure in that sense. If the central bank thinks it’s achieved its easing objective and at some point reverses into tightening, the paper inflation rate starts getting knocked back toward the silicon inflation rate, which at some point may be declining itself. There’s a sort of a convexity or leveraged effect in the way the two rates behave together according to the cycles of easing and tightening through negative rates and paper depreciation.

One reason why a silicon medium of account makes sense to me is that the banking system should naturally be denominated in silicon money rather than paper money, since it is the conveyor of silicon interest rates.

(I wonder – do monetary economists ever differentiate between the medium of account for goods and services prices versus what is used in accounting for financial assets? I guess in the normal course there is no reason to do so.)

Finally, it would be interesting to work through the logic of why a managed depreciation of paper would be recommended along with negative rates, while a managed appreciation would presumably not be considered with positive rates. In other words, why would there be a zero upper bound with respect to the limit for any marginal appreciation of paper money (as it might occur when interest rates get reversed from (1) per cent back to 0 per cent, for example). I suspect there’s an easy and/or interesting answer to this, but I’m feeling too lazy to think about it right now.

TMF 3:14

yes

JKH said: "Suppose at first the medium of account is silicon.

And the unit of account is $ 100 silicon.

Suppose the central bank sets interest rates on silicon at (1) per cent."

Is that interest rate on IOR or on the fed funds rate?

Pardon my comment above, silicone implants, false accusations and all. I wrote it without looking at the post (which I'd read some days before) only some comments. One question, though:

Why is the medium of account so important? Aren't there many economies with different media of account that operate side by side? Dollarised nations, e.g.. Which is the more important inflation rate in such a system? And why?

Or is that exactly what you're trying to say, e.g. when you write: There is no difference, except in the names of things, between a bank with a fixed exchange rate paying 1% interest and a bank with a 1% appreciating crawling peg paying 0% interest.?

In which case I think I tentatively agree with your post.

Having said that, my beef seems to (always) be your contention that alpha money in one form or the other is per definition 'salient'. But my point is, I guess, that your post works just as well if you drop that premise.

Everyone: maybe switch the discussion to my new post, where (I think) I have clarified/re-drawn the policy implications, by starting with a different question.

The comments to this entry are closed.

Search this site

  • Google

    WWW
    worthwhile.typepad.com
Blog powered by Typepad