[cboyack]

Prof. Herbener,

I was recently in a debate with some people regarding the specific mechanics of quantitative easing, especially when it comes to the purchase of Treasuries. The specific question is: in QE, did the Fed pay the banks for the Treasuries, or did the money go to the Treasury?

Let’s place to one side QE that involves the Fed purchasing mortgage backed securities. That is clearly the Fed exchanging cash (printed from thin air) for assets held by the banks — assets of questionable value.

But is the scenario different when it comes to that part of QE in which the Fed buys Treasuries?

Specifically, I was debating someone who claimed that the Fed was essentially buying Treasuries directly from the Treasury, and not from banks (or brokers in the business of selling bonds). His claim was, essentially, that the bulk of the trillions that were created in QE went to fund government programs, like welfare programs, unemployment insurance, and the like. To the extent that the Fed was buying Treasuries from financial institutions, he claimed, the banks (or brokers) served as a “pass-through” — with the banks purchasing the securities only momentarily, and immediately re-selling them to the Fed, to allow the Fed to evade laws against buying Treasuries directly from the U.S. Government.

He admits that the money ended up as bank reserves, but he believes that it first financed all manner of government programs before making its way to the banks.

Based on the reading I have done, I espoused a very different view: that the Treasuries purchased in QE had already been held by banks for some time. That the idea of QE was to replace financial instruments (like mortgage backed securities and long-term Treasuries) with cash, to serve as a “credit easing” that would loosen credit markets and boost lending. From what I can tell, the money went straight to bank reserves, and (for the most part) has not made it out — especially given the Fed’s policy of paying 25 basis points on these millions/billions/trillions.

Sure, the Treasuries were at one point purchased from the Treasury, giving money to the federal government. But that happened when the banks bought them. When QE happened, though, in my view, it was strictly a transaction between the Fed and banks. The Fed created money, and put it in bank reserves, and banks in exchange gave Treasuries they owned to the Fed.

So: which one of us is right? Thanks for your time.

[jmherbener]

Here is a Fed paper explaining QE. The authors note that the purchases of Treasuries were conducted by the FRDB of New York in the same manner as it conducts regular Open Market Operations (pp. 9-10). The New York Fed bank holds auctions for primary dealers, who sell the Treasuries to the Fed.

http://www.ny.frb.org/research/staff_reports/sr441.pdf

The article notes that the New York Fed Bank posted a summary of its QE purchases on its website as the program went along (p. 10, note 17).

http://www.newyorkfed.org/markets/pomo/operations/search.html#

To determine whether or not the Fed was shrinking the time between the sale of the Treasuries by the Treasury and the purchase of the Treasuries by the Fed, one would have to compare the normal OMO purchases before 2008 with the QE purchases. Given the Fed article’s description of the program, there doesn’t seem to be any evidence that the time lag shrank or, at least, that the Fed purchases were immediately after the Treasury sales as your friend claims. (One could go through each purchase on the website to see if the Fed made any immediate purchases.) In any case, your friend is correct that the Fed purchases of Treasuries, whatever the lag happens to be, provide lower-interest-rate-than-otherwise funding for the Treasury, which spends the funds on general programs of the Federal government. But the Fed indirectly funding the Treasury expenditures in this way is always the case for all Fed purchases of Treasuries and not a special consequence of the QEs.

[Author]

Posts