Tuesday, November 6, 2012

Miles Kimball on Electronic Money

Miles Kimball has an article on Quartz about electronic money.   He advocates making electronic money the medium of account, and allowing tangible paper currency to trade at a discount during national economic emergencies.

He proposes that households and firms have electronic money created by the government available, though he doesn't really spell that out.   He mentions in passing that the interest rate paid on that electronic money (effectively "reserve balances" held by firms and households at the central bank) should be set below the interest rate target set by the Fed.    Perhaps I was reading between the lines, but he seems to be advocating a system of using the interest rate paid on the banks' reserve balances as the instrument of monetary policy.   His major emphasis is that this target interest rate could be reduced below zero if the Fed found it necessary to do so in order to maintain spending on output.

As for tangible hand-to-hand currency, it would trade at a growing discount relative to the electronic money held by households and firms.   While Kimball doesn't explain, I suppose he is assuming that by reducing the price at which currency can be deposited in exchange for electronic money in the future, its actual market value would fall.   (I think this is likely true.)

This discount on tangible currency would only occur in national economic emergencies--that is--when the interest rate that banks earn on reserve balances is reduced to very low levels.   Again, while Kimball wasn't too clear, when the interest rate the Fed pays to banks rises enough that the lower one paid to firms and households rises enough, so that the still lower rate of price decrease on the tangible currency hits zero, then tangible currencywould  no long fall in value.    When good times return, he explains, the discount would be reduced (the price of currency would rise again) creating a positive rate of interest, presumably still slightly lower than that on electronic money.  The point is to make tangible currency perform slightly worse than a balance in an electronic money account.

In my view, there is no need to set up deposit accounts for households and firms at the central bank (Kimball's "electronic money.")   The private banking system already provides those--they are called checkable deposits.     It would be possible for the government to adjust the rate of exchange between reserve balances and currency at the central bank and allow private banks to use that same rate of exchange for withdrawals and deposits of the tangible currency.    The key, however, isn't how much currency can be withdrawn from deposits, but rather the price at which it will be accepted for deposit in the future.   A dollar withdrawn when interest rates turn negative must be only accepted for deposit when interest rates turn positive again at a lower price reflecting the negative interest rate on deposits during the interim.

I still believe that rather than have the government develop a new sort of money for households and firms  (electronic deposits held directly at the central bank,) it would be better to leave the issue of the tangible currency to the private sector.

 In my view, a situation where extremely low (and even negative) yields on short and safe assets makes currency more attractive to hold, is not a time to start reducing the price of currency.    The growing discount on currency means that each dollar-dominated deposit is being redeemed with progressively more hand to hand currency.   While that make it more attractive to hold deposits, (with the nominal interest rate calculated in terms of the currency remaining approximately zero,) for this to all work out, the government must stick to its commitment to only allow the currency to be redeposited at that lower price.    Or rather, to only gradually raise the price at which it can be deposited at a rate reflecting interest rates on other short and safe assets when they are again positive.   Allowing that process to be short circuited, and having the currency jump back to par would be a problem.  And why wouldn't everyone in the private sector clamor for such a jump in the value of  their currency holdings?

In my view, private hand-to-hand currency, where banks just cease issuing it when it is no longer profitable, and have a call provision when maintaining the outstanding balance is too costly, is a much better system.   Rather than trying to reduce the price of currency, the central bank just sets the interest rate it pays on deposits where it thinks it is best.   Banks set interest rates on "electronic money" as they think is best.   And if no one wants to issue currency, then none is issued.

Sure, the lack of hand-to-hand currency is inconvenient.   Let the private sector come up with substitutes to solve the problem of currency shortages.   I think it is simpler than managing a currency deposit exchange rate.


  1. Bill,

    I assume you (and Miles) mean that hand-to-hand currency will trade at a premium, not discount. It has a higher coupon.


  2. Read the article. I might be wrong, but I read it as below:

    I think that the deposit price of currency is supposed to drop at a rate slightly more than the negative nominal interest rate on electronic money.

    Next week, a one dollar bill has a price of .99. Then it is .98, and so on.

    You can buy a paper dollar from your bank at that price. You can deposit that amount and get a dollar balance in you deposit account.

    The rate of capital loss on currency would equal the negative nominal interest rate on electronic money.

    When interest rates turn positive again, then the price of currency would rise at a slightly lower rate than the positive interest rate.

    This week is is 78 cents per dollar bill. Then it is 79 cents. Eventually you get back to $1 per dollar bill.

    The face value of the currency is constant. The value in terms of deposits falls and then rises.

    A growing and then shrinking discount.

    It is peculiar. Everyone wants currency because it is a great store of wealth when interest rates are negative. And rather than the price of currency rising relative to deposits, it falls.

    What drives this is that the currency will only be accepted for deposit at the falling and then rising price.

  3. Not redeeming the existing FRNs at face value would be legally problematic - it would be a kind of default.

    But I suppose you could create a brand new currency that works this way, with a dollar exchange rate which *permanently* changes every day at the rate of interest (forget about "returning" to 1:1, just let it go whereever it goes!)

    1. It is a bit odd to talk about redeeming FRN at any value. The status quo is that deposits are redeemed with FRN at face value.

      When a discount is placed on FRN, then extra FRN are being provided. To say that you can buy a $1 FRN at your bank for a 99 cent deposit means than you can withdrawal about $1.01 in currency for a $1 balance. Hardly default on the deposits.

      And FRN aren't redeemable into anything now.

      Still, the point of the proposal is to start thinking in terms of deposits, and so your $1 FRN won't buy a $1 deposit, but only a 99 cent deposit.

    2. I didn't say it would be a default on deposits. It would be a default on the FRNs. Yes, you can default on currency! "This note is legal tender for all debts, public and and private." It doesn't say "at face value", but that is obviously implied.

  4. Bill,

    I did read it, but I don't get it. You can see my comment to Myles on his blog. Basically I don't understand what monetary operations he is proposing, and I don't see what he can do to control the value of hand-to-hand currency. Without stamping it (negative interest) I don't see why it would trade at a discount. If paper money trades at a discount I'll buy it in unlimited quantities by borrowing at negative rates and make big arbitrage profits. It has to trade at a premium equal to the discount factor (greater than 1) up to the time when rates go positive again. What am I missing?

    1. The key is the redeposit value.

      First, think about a $1 deposit. If you do noting with it, when interest is negative, it gradually drops in value as negative interest is deducted. Say, after 2 years, it is down to 96 cents. Then, for a time, the interest rate is no longer zero, but zero. And so the balance stays 96 cents. Then, the interest rates become positive again. The balance starts rising. Suppose it is one percent. After about 4 years, it is back to $1. Of course, if you leave the money in the bank, it would be $1.01 after another year, but forget that part.

      Now, we make the dollar value of currency do the same. The currency has a $1 face value (it looks like a FRN.) During the period interest rates are negative, the price drops 2% a year. After 2 years, the price is 96 cents. Then, when interest rates are zero, it stays 96 cents. Then when interest rates are positive 1 percent, the price of the currency rises gradually. After 4 years, it is back to $1, the face value.

      Currency had a growing discount, reaching 4%. It stayed steady for a time, and then the discount gadually falls over time-- 3%, then 2%, and after 4 years it is 0%, no discount at all.

      What allows this to happen is how much currency is accepted for deposit into bank accounts.

      I take out one dollar today. I have a $1 FRN. My balance at the bank is down $1. The price of the currency is dropping at 2% per year. What that means is that after 6 months, if I choose to deposit the dollar back into the account, I don't get a dollar balance, I get a 99.5 cent balance. If I wait one year, it will be a 98 cent balance that I get for depositing my $1 denominated Federal Reserve note. After two years of leaving the FRN in my sock drawer, if I deposit it, I get a 96 cent balance. (I am ignoring compounding.)

      Now, if I wait long enough, then eventually my $1 FRN in the sock drawer will give me a $1 deposit. But if I had a one dollar balance in my bank account all along, it would eventually return to $1. It would decrease as long as interest rates are negative and then increase when they turn positive again.

  5. K, "discount" doesn't have any meaning if the money is not redeemable at face. You can't buy a $1 bill for $0.99 of base money and then convert it to $1 of base money. You can only convert it back to $0.99 of base money. It's a permanent change to the currency/base exchange rate.

  6. Another great post! Hope you don't mind me linking to your blog at fifthestate.co

  7. Max,

    I never said you could do that.

    Assume the Fed is at 0% and then announces it's going to cut the policy rate to -5% for two years and then back to non-negative rates. So I borrow one e-dollar and convert it at par to a paper dollar. I hold the position for two years, collecting 5% interest per year, then convert my paper dollar back to an e-dollar (at par again) and repay my loan.

    You can't just set rates *and* the price of paper dollars (bonds) where ever you want without permitting arbitrage. The price of paper dollars has to be exp(-rT) where T is the time at which rates return to zero, and r is the (negative) yield up to that date.

    The example holds whether the conversion at the beginning and the end is par or 0.99. It has nothing to do with it. I make 10% interest for free over the two year period.

    1. When you pay back the loan, each FRN in 2 years will be worth about 90 cents.

      Borrowing FRNs and holding them would provide no benefit.

      By the way, rate of price decrease in FRN's is supposed to be lower than the interest rate at which they can be borrowed.

      I think Kimballs notion would be that the Fed begins debiting banks reserve balances at 5% a year. Members of the general public who want to have e accoutns at the Fed have those debited by 5.2 percent each year. And the price of Federal reserve notes fall a 5.4% each year.

      If you can get a bank to lend you a dollar at -4%, then you could withdraw a $1 FRN. It's price would be dropping at a 5.4% rate. After a year, its price would be 94.6 cents. You could deposit it for a 94.6 cent e account balance or a 94.6 balance in your checking account. Then, you would owe your banker 96 cents.

      You have lost 1.6 cents.

      When interest rates are back in positive territory, say the policy rate is .5 percent, then the price of the Federal Reserve notes would start to rise from the current, below par, values at a .1% annual rate. Eventually, they return to par.

      When the policy rate is back up to 5% (positive 5%,) then the FRS price would be rising at 4.6 percent per year. Before long, they would be back to face value. And then they are left alone until next time the policy rate gets below .4%.

  8. "I hold the position for two years, collecting 5% interest per year, then convert my paper dollar back to an e-dollar (at par again) and repay my loan."

    You can't convert back at par. There is no par. So you are losing more than 5% a year on the paper currency, and you never get it back.

    Again, this couldn't be applied to existing FRNs because that would be a default. It would be a completely new currency where the face value means nothing in terms of dollars...although the conversion rate could, sometimes, be 1.0, purely as a mathematical convenience (not because it's a dollar bond!)

    1. Default of what?

      Who is obligated to accept FRN for deposit at par?

      Par means that the price is equal to the face value. Discount means the price less less than the face value. Premium means the price is more than the face value.

      For Kimball, these prices are in terms of deposits. Par means that a $1 deposit is exchange both ways with a Federal Reserve note denominated as $1. Premium means that $1 Federal Reserve note is exchange both ways with a deposit of more than $1. Discount means that a $1 Federal Reserve note is exchanged both ways with a deposit of less than $1.

      In the olden days, premium, discount, and par would refer to gold. Redeemability would keep FRN at par with gold. 10 $1 Federal reserve notes would be at par with a $10 gold eagle. If the Federal Reserve notes were at a discount, for example, if redeemability had been suspended, then each FRN would would be at a discount relative to gold. In practice, what happened in similar circumstances, is that gold coins would trade at a premium relative to their face values.

      Anyway, with a gold standard, a discount on paper money is a kind of default. And it happened quite frequently. But rather than liquidation, contractionary credit policies would be followed until the paper money rose back to par.

      But there is no gold standard and FRNs aren't redeemable for anything. On the contrary, deposits are redeemable for FRNs, and no part of the proposal involves anyone redeeming a $1 deposit with anything less than a $1 Federal Reserve note.

    2. If you owe somebody $100, and the government won't redeem your FRNs at face value, then a $100 FRN won't settle the debt. That's a default on the FRNs.

      You are only looking at it from a deposit holder point of view. I agree there's no default on deposits.

  9. Incidentally, since this scheme allows currency to earn interest (at any rate less than or equal to Fed Funds), it means that Milton Friedman's "optimal quantity of money" can be achieved at any inflation rate (rather than requiring deflation at the Fed Funds rate).

    If that matters. :-)

  10. Bill, Max,

    I get it. I had assumed that paper money would return to par when rates were positive. It doesn't. Here's the key bit of Miles' paper:

    "In a recession, this would mean that the discount for paper dollars would gradually widen, but in good times (when real interest rates tend to be higher) the discount would narrow until the paper dollar was again at par with electronic dollars, where it would stay until the next recession."

    The way I would put it is the log exchange rate is given by the integral of the spot rate from the time the spot rate first goes negative until the *integral* returns to zero. At all other time the log exchange rate is zero (exchange rate is one). The derivative of the log-price (the profit rate of holding dollars) during periods of non-par exchange is r, as is required for all risk free assets.

  11. I'm with Max. If I have a $1 debt, I have the right to extinguish it with a $1 bill. It says so, right there on the bill. I think you'd first have to redeem those bills for new ones with a little asterisk beside the value, and a whole new disclaimer. I don't see why you couldn't do that though. Germans had to give up their Bunds for Euros whether they liked it or not.