Friday, January 11, 2013


If you have not run across this, you have probably been hibernating for the last few weeks. Do a search on #mintthecoin, and fill yourself in on the issue at hand. Where political conflict meets crackpot monetary economics, there is plenty of room for nonsense, of course.

The closest historical parallel to this I could find (and you may know of others) was the issue of playing-card money in New France, beginning in the 17th century. I good account is in this Bank of Canada publication. Typical practice in the colony in New France was for the King of France to make gold shipments to the colony, which were then used for government expenditures in the colony - mainly paying the troops. In 1685, the shipment was delayed. The troops had to be paid, and an inventive French official in the colony thought up a scheme for getting this done. Playing cards in the colony were requisitioned, autographed, made out as IOUs in varying denominations, and used to make government payments. These notes were promises to pay gold in the future, and they circulated as media of exchange. The first playing card issue was in fact subsequently retired, though the colony would repeatedly resort to this expedient in the future, and the promises eventually became weaker.

It's not like this went unnoticed in France, though. The colonials got some resistance from the King of France:
He [His Majesty] strongly disapproved of the
expedient which he [de Meulles] has employed of circulating card notes, instead of money, that being extremely dangerous, nothing being easier to counterfeit than this sort of money.

You can see how this is like the #mintthecoin controversy. Faced with the intransigence (or tardiness) of another branch of government, new liabilities are invented that will allow the government to conduct its business, and not everyone is happy as a result.

Private financial intermediaries are also in the business of inventing new financial instruments, in order to sidestep regulations, or to take advantage of new technologies. An interesting historical example, which actually shares some of the features of the New France playing-card money episode, was the use of clearinghouse certificates in pre-Federal Reserve banking panics in the United States (see this paper by Gary Gorton). It's useful to think of the banking panic episodes that occurred in the period 1863-1913 as currency shortages, typically set off by the failure of some large financial institution(s). Depositors, in a panic over what bank might fail next, ran on banks to convert deposits into National Bank notes (circulating paper money backed by government debt) and gold. The suspension of convertibility by banks would then result in significant disruption of payments.

Later in the National Banking era (1863-1913), clearinghouses thought up a means to mitigate the panics. For example, in the 1907 panic, banks in the New York clearinghouse suspended convertibility of deposits into gold, but were authorized to issue clearinghouse certificates, which circulated as media of exchange and could be redeemed after the panic was over. This appears to violate the rules governing banks in the United States at the time, as National Banks could issue National Bank notes only under certain conditions (principally, the notes had to be backed - essentially in full - by federal government bonds), and state-chartered banks faced a punitive tax on note issues. Presumably a clearinghouse certificate was a liability of the clearinghouse and not of an individual bank so as to skirt banking regulations on note issue.

In contrast to the experience with clearinghouse certificates, private banking innovation can of course sometimes be detrimental to social welfare. Part of the narrative of the financial crisis involves the use of off-balance sheet vehicles - structured investment vehicles and special purpose vehicles, for example - by financial institutions. Certainly these activities were not benign, and it is an open question how such entities should be regulated.

What about #mintthecoin? Most of the elements in the #minthecoin story are contained in the above anecdotes. There is financial innovation. That innovation involves moving some activity off a balance sheet so it will not cause trouble. Engaging in this new activity could prevent a major disruption of payments, but could also cause some other trouble.

As has become more clear in the last few years, monetary and fiscal policy are closely intertwined. It's hard to tell where fiscal policy stops and monetary policy starts. When the Fed swaps short-maturity government debt for long-maturity government debt, or purchases mortgage-backed securities with reserves, are those monetary policy actions or fiscal policy actions? You tell me.

Broadly, here's what the Treasury and the Fed do in the United States. In order to finance its deficit, the Treasury issues interest-bearing debt in various maturities. The Fed then engages in what amounts to debt management. It can purchase the interest-bearing debt of the Treasury and issue other types of liabilities - outside money - which consists of reserves (essentially transactions accounts of financial institutions with the Fed) and circulating currency. What is ultimately important is the debt of the consolidated Treasury and Fed - its size and composition. Both size and composition are determined jointly by the actions of the Fed and the Treasury. Ultimately, the outstanding consolidated debt consists of interest-bearing securities, reserves (now interest-bearing as well), and currency (not interest-bearing).

Legislation puts some restrictions on how the Fed and the Treasury interact, and what kinds of liabilities they can issue. The Fed cannot lend directly to the Treasury - the Fed buys government debt on the secondhand market. The Treasury cannot issue Federal Reserve notes or bank reserves. The Fed cannot issue marketable interest-bearing securities.

There are loopholes, however. Under the Federal Reserve Act, the circulation of Federal Reserve notes falls under the Fed's purview. But where and how are Federal Reserve notes produced? By the Bureau of Engraving and Printing, which is a government agency. It's actually part of the Treasury Department. Similarly, coins, which constitute part of the stock of currency in circulation, are produced by the U.S Mint, also a government agency and part of the Treasury Department.

Reserve accounts at the Fed are essentially transactions accounts for financial institutions, used for interbank clearing and settlement. These reserve accounts have withdrawal privileges. For example, a bank with a reserve account can make withdrawals of Federal reserve notes and coins from its account, and can make deposits of notes and coins. Part of the business of regional Federal Reserve banks is to manage the stock of currency. Regional Feds take delivery of notes and coins from the printer and the mint, and in general maintain the stock of currency and manage its growth. Presumably there is some accounting transaction when the printer or the mint delivers new currency to the Fed, but that transaction is irrelevant, as ultimately any profit the Fed makes is returned as a transfer to the Treasury.

Until recently, like me you were probably under the illusion that the Treasury could not issue outside money. Not so, apparently. The Treasury is allowed to issue special coins, and to put them directly into circulation. This is what #mintthecoin is all about. Typically, the innovation is described as the issue of a $1 trillion coin by the Treasury, which is deposited with the Fed. The Treasury then spends out of its account with the Fed, and thus sidesteps the Federal debt ceiling, which applies only to the stock of interest-bearing Treasury debt.

What's that about? Many non-bank institutions hold reserve accounts with the Fed. These institutions include American Express, the GSEs (Fannie Mae and Freddie Mac), and the U.S. Treasury. The Treasury account with the Fed that is currently in use is the General Account. Since the financial crisis, the balance in this account has varied substantially, but averages about $60 billion. Presumably the Treasury general account carries with it the same privileges as any other reserve account, though I assume it does not bear interest (not that this would matter), as the GSEs do not get interest on their reserve balances. Thus, there appears to be nothing that prevents the Treasury from minting coins, depositing them in its reserve account, and spending out of the account.

As far as I know, no one has pointed out that the Treasury could also mint coins and deposit them in private depository institutions. Through the Treasury Tax Loan Program, which was more relevant before the financial crisis (when the management of the stock of reserves was more critical), the Treasury holds accounts at various financial institutions. Typically, these accounts were used for deposits of tax revenue, but there seems no good reason why the Treasury could not deposit coins in such accounts.

In terms of accounting and standard fiscal/monetary policy, there is nothing weird about #mintthecoin, as some other writers have stated. Whether the coins are deposited directly with the Fed, or with a private depository institution, ultimately they will end up at the Fed. If the Fed were to respond passively, the stock of outside money goes up by the amount of the Treasury coin issue. This is identical to what happens when the Treasury issues debt and the Fed buys it. The Fed need not be passive, however. One possibility is that the Fed sells an amount of government debt from its portfolio equal to the coin issue. Under current circumstances, given that the entire stock of debt held by the Fed is long-term, the net effect is identical to what would happen if the Treasury issued long-term debt instead of coins. And under current circumstances, if the Fed is passive, this is equivalent to what would happen if the Treasury issued short-term debt, as currently short-term debt and reserves are essentially identical assets.

Surely we can be more inventive than #mintthecoin, though. Here are two other suggestions. I'll leave it to you to determine the legality of these options, or who might get upset.

1) The off-balance-sheet option: Fannie Mae became a private institution, and Freddie Mac was established, as part of a Johnson administration move to take the mortgage market activities of Fannie Mae off the federal government's balance sheet. Suppose that Fannie Mae were to issue agency securities and use the proceeds to pay salaries at the Pentagon, or Pentagon employees were to temporarily become employees of Fannie Mae. Currently under government "conservatorship" Fannie Mae has to do what the federal government tells it to do, but its agency securities are not part of the government debt for accounting purposes.

2) The playing-card or clearinghouse certificate option: Federal government departments could issue their employees certificates promising payment in the future, in lieu of salary. I'm not sure what makes a government debt obligation fall under the debt-ceiling limit, but the idea would be to design the IOUs so that they don't meet those criteria. Like the playing cards and clearinghouse certificates mentioned above, there's nothing to stop people accepting these IOUs in exchange.

What should the President do? If I were him, I would try to stay out of trouble. The federal debt limit is idiotic, but we're stuck with it. The President is playing a repeated game with incomplete information. The key piece of private information concerns the payoffs for his opponent (and maybe for him as well). I'm not a game theorist, but I think the relevant theory tells you that anything can happen, so I think Barack is on his own on this one.


  1. Good post. Like the historical flourishes.

    Bernanke could always refuse to accept the coin. The Federal Reserve Act doesn't obligate him to accept deposits from the government, he "may" accept them. Bernanke is required to protect his balance sheet by properly collateralizing the note issue. A $1,500 coin hardly collateralizes $1 trillion in new reserves. Lastly, legal tender can be refused in transactions - it's only necessary to accept it in the settlement of debts, never in spot.

    "2) The playing-card or clearinghouse certificate option: Federal government departments could issue their employees certificates promising payment in the future, in lieu of salary."

    Wow, you're agreeing with Krugman. See Moral Obligation Coupons.

    Not sure that IOUs would qualify as legal tender, so government creditors wouldn't *have* to accept them. That might hurt their ability to get off the ground.

  2. Very nice post, thanks! Do you think there is any chance we can get rid of the idiotic debt limit altogether?

  3. In a consolidated Fed/Treasury balance sheet, Fed actions only affect the duration of gov't liabilities. T-bond QE shortens the duration considerably since ER's have effectively overnight duration.

    At the extreme, the gov't is funded by o/n duration ER's. This balance sheet is highly exposed to increases in real interest rates.

    Such a huge public-sector duration bet would signal to markets that the "consolidated" entity would not stomach the higher tax liabilities (duration losses) generated by positive real rates. The result is a Latin American-style real rate-velocity feedback loop: facing persistent future negative real rates, agents hedge against real wealth erosion, causing prices to rise and real rates to drop further, causing more hedging, etc.

    1. "This balance sheet is highly exposed to increases in real interest rates."

      If the (consolidated) government was only concerned with the cost of servicing debt, on average they would want to borrow short - based on the observation that the yield curve is typically upward-sloping. However, currently the government could borrow for 30 years at a lower rate than the average 1 month T-bill rate we have observed for the last 20 years, so this would seem like a good time to increase the maturity of the outstanding government debt, rather than reducing it.

      I'm not sure though, as the government's goal isn't simply to minimize debt service costs. What are the consequences for the sharing of aggregate risk of the composition of the government debt? Economists have not thought enough about this problem for us to give good answers.

  4. Imagine an economy with no credit risk -- only duration risk. That risk accrues to intermediaries engaging in maturity transformation. The question then becomes frictions are created when those intermediaries become insolvent as a result of (real) interest rate movements.

    At one extreme, the banking system holds all the duration risk. If capital is insufficient to cover potential losses, then depositors will seek alternative assets: mainly currency. This is deflationary.

    At the other extreme, the consolidated gov't balance sheet holds all the duration risk. The balance sheet's "capital" is the fiscal authority's willingness/ability to raise future tax revenues. If that "capital" is insufficient to cover losses, "depositors" will seek alternatives: goods, assets, and other currencies. This is inflationary.

  5. Thanks for the post which provides useful information. Recent debate reminds me 2011 debt ceiling crisis. I think "mint the coin" will only delay the problem rather solving it. Given that the Treasury spends $100bn per month, issuance of $1tr averts the FED debt ceiling only for a year or so. In fact, we will exactly be at the same point with no solution as we were in 2011. Plus, it would lead the inflation to undesirable levels. That's why I think "mint the coin" is not a reasonable strategy to mitigate the debt ceiling problem.

  6. Imagine an economy with no credit risk -- only duration risk. That risk accrues to intermediaries engaging in maturity transformation. The question then becomes frictions are created when those intermediaries become insolvent as a result of (real) interest rate movements. cheap nfl football jerseys
    Cheap Nike NFL Jerseys USA