Wednesday, October 27, 2010


The FOMC meets next week, and some form of quantitative easing (QE2) program is anticipated. This Wall Street Journal article reflects the view that this will be a "measured" approach, i.e. "a few hundred billion dollars" in purchases of 2 to 10-year Treasury securities over several months. The article quotes publicly available sources - mainly published speeches by Bernanke and Fed Presidents, so there is not much in the way of specific new information.

The "measured" approach is consistent with remarks made by Jim Bullard, in post-dinner talk at this conference. Jim is on record here as being concerned about deflation, with QE2 part of the remedy. In his remarks last week, he said (and you have to trust my memory here) that, in his opinion, quantitative easing should proceed much as traditional FOMC decisions about fed funds rate targets have. He thought QE should proceed a bit at a time, with re-evaluation of the program at each FOMC meeting, which certainly seems to be consistent with what the WSJ is anticipating.

Given that the Fed is going to do QE2, I think the gradualist approach is appropriate. We have no well-established theory for understanding what the effects of QE2 will be (though Kocherlakota does a nice job here of laying out the issues), and no practical experience that will help us forecast the quantitative effects. Further, I think there is significant upside inflation risk (though I know a lot of people don't agree with me). The stock of currency is increasing at a higher rate recently, the US dollar is depreciating, and some commodity prices are increasing at a high rate. Of course, some of what we are seeing (for example higher "break-even" inflation rates implicit in bond yields) can be attributed to anticipated easing by the Fed, but I think the Fed would be correct in being concerned about the potential for more inflation than what we would like.


  1. here's what I just don't get. If the Fed can raise the interest rate paid on reserves, raise statutory reserve requirements as high as it likes in conjuction with simply selling some of it's holdings why would they ever lose control of inflation on the upside?

    Now, there is the ubiquitous question of political will to tighten but that is there anyway. If they don't have the will to tighten to prevent an overshoot then this can happen regardless what they buy today.

    Furthermore, if the Fed could lower inflation but chooses not to then by definintion that inflation is not more than they want.

    So again, why would inflation be uncontrolable to the upside? No matter how much they buy.

  2. 1. I don't think you want to think about using reserve requirements as a policy tool. That should not be on the table.

    2. You're right that, in principle, the Fed can increase the interest rate on reserves and get all the inflation control they want, but I am thinking through the "political will" problem. Acquiring a portfolio of long-term debt implies commitment, because of the capital loss the Fed inflicts on itself if it does raise the interest rate on reserves. The more long Treasuries they acquire, the bigger the potential loss if they tighten down the road, and therefore the less likely they are to do it.

  3. A higher interest rate on reserves would definitely push the short-end of the yield curve up, and probably also the long-end. Not good for a fragile recovery, and I'm not sure the Fed has the "political will" to choke the recovery.


  4. I'm really completely changing the subject here, but have you seen this new paper on fiscal multipliers?

    Basically floating exchange rates, trade interconnectedness, "developing" status, and high indebtedness result in tiny or zero multipliers. It's almost like resurrecting an idea that had been academically dead in the water for 30 years or so was dumb or something.

    Link courtesy of Marginal Revolution

  5. No, I hadn't seen this. What do you think of VAR results? It seems the people who use these want to argue that they are agnostic, but I have a hard time believing the identification.