Saturday, February 13, 2010

The Treasury Prints Bonds Like the Fed Prints Cash

In the wake of Warren Mosler's recent interview on CNBC's Squawk Box, I thought it made sense to clarify a point about government "borrowing." I put the word borrowing in quotes of course because the government is not really borrowing money when it issues Treasury securities. Mosler tried to explain this by running through the way Federal Reserve accounting worked, but I am afraid that most people's eyes glazed over. Certainly Becky Quick at CNBC had no idea what he was talking about.

Generally, people understand that the government retains the power (through the Federal Reserve) to print as much as cash as it wants to buy Treasury bonds and therefore to monetize the public debt. What is not widely understood, though, is that the Treasury itself can spend as much money as it wants without forcing the Fed to do anything at all. We can have laws that arbitrarily prevent the Treasury from printing cash (and we do to a certain extent), but that still does not constrain the Treasury from spending an unlimited amount of money.

How can that be if the Treasury has to raise the cash first from tax revenues or the issuance of Treasury bonds? What if investors, or the Chinese in particular, decide not to buy new Treasury bonds? Well, let's think about what happens when the Treasury's Federal Reserve account is completely depleted, and it wants to spend an additional dollar:

It has an auction for $1 worth of Treasury securities. An investor, usually a bank which has extra cash sitting around in danger of earning 0% interest overnight, bids some positive interest rate in the auction and buys it.

The investor has $1 worth of Treasury securities, and the investor's bank account is debited by $1, which means also that his bank's account at the Federal Reserve is debited by $1 (or, if it is a smaller bank which does not have a Reserve account, there is a debit in that bank's account at a bigger bank, etc.). And the Treasury has $1 credited to its Reserve account.

The Treasury now spends that $1 by transferring it to the bank account of whoever the recipient is (whether it be a social security beneficiary, a doctor who performed Medicare services, a construction company working on a stimulus project, or a bribe to the head of a union who promises to "get out the vote" for a Democratic candidate).

Now, that dollar looks like a credit in the recipient's bank account, but it is also (following the upward chain of accounts) a credit to a big bank's account at the Federal Reserve.

So the total amount of bank reserves at the Fed remains the same. All that has happened is that the total amount of Treasury securities has increased by $1.

Note that the analysis is unchanged if the recipient of the Treasury's largess decides to spend or invest his windfall. Whoever he buys goods or services or stocks from will now have the credit in his bank account and therefore in his bank's Federal Reserve account.

It should be clear by now that the money that the Treasury spent has essentially gone to fund the purchase of the Treasury security which the Treasury originally issued to raise the money to spend in the first place.

It is a perpetual motion machine, and by induction we can see that the Treasury can spend any amount of money and the accounting effect of that spending is that the Treasury prints an amount of Treasury securities equal to the amount of its deficit spending.

We have argued in the past that a Treasury security is not fundamentally different from cash. It is a government IOU, same as cash, except that it earns a positive interest rate rather than no interest. It's true that you can't spend a Treasury bond as easily as cash, but the Federal Reserve stands ready to lend to any member bank an unlimited amount of cash against an equivalent market value of Treasury bonds at slightly below the Fed's target interest rates, so the link between Treasuries and spendable cash is very strong.

So the Treasury's ability to print bonds is equivalent to the ability to print money. By itself, it can increase private sector net financial wealth.

3 comments:

Anonymous said...


It should be clear by now that the money that the Treasury spent has essentially gone to fund the purchase of the Treasury security which the Treasury originally issued to raise the money to spend in the first place.


That's a bit too confusing a sentence to be "clear", but I think I understand it.

Can you talk a bit about what determines the shape of the yield curve? Or to put it a different way, how does the Treasury decide at what durations to borrow money?

-DWJ

ESM said...

LOL. The claim is that the idea is clear. Not that the sentence is clear (in the sense that it is easy to parse). Actually, I intentionally made the sentence have a circular structure as a metaphor for the way the money goes around in a circle. You can see why I don't write more often...

The shape of the yield curve of course is driven by the supply and demand for loans of varying maturities. The government does have a big effect on the yield curve because it issues a lot of longer-term bonds relative to the size of the corporate and mortgage bond markets. I like to think of the government's effect in terms of how much duration risk it issues and not in terms of absolute dollars it is "borrowing" through the issuance of bonds. Since the money is going around in a circle, what really matters is how much interest rate risk the Treasury is asking the private sector to absorb.

Now, I have no idea why the Treasury asks the private sector to absorb any interest rate risk at all. As Warren Mosler has pointed out, there is no reason for the government to issue anything other than 1-day interest-bearing deposits at the Fed (the equivalent of 1-day T-bills). The yield curve can be filled out well enough by the private sector through the trading of longer-term corporate bonds.

I suppose one could argue that there is a public good in providing a "public option" for investing long-term. That is, perhaps it is beneficial to give very risk-averse entities a way of locking in 30-year returns in a credit risk-free investment by offering 30-year fixed rate Treasury bonds from time to time. But there is certainly no Treasury funding purpose that makes sense.

As to how the Treasury debt management people make their decisions regarding how many bonds to issue of which type, and when, I have almost no idea. I suspect that very little intelligent thought goes it to the process.

ESM

Anonymous said...

I suppose one could argue that there is a public good in providing a "public option" for investing long-term. That is, perhaps it is beneficial to give very risk-averse entities a way of locking in 30-year returns in a credit risk-free investment by offering 30-year fixed rate Treasury bonds from time to time. But there is certainly no Treasury funding purpose that makes sense.


Right, I seem to remember reading a while back (though I can't find a link) that life insurance companies bid up the price of longer-term Treasury debt to offset some of their liabilities.

If that's the case, then the Treasury isn't really providing a public good as much as it's getting a better price for issuing longer-term debt, right?

-DWJ