Progress Report on QE2

We’re now 3 months into the Fed’s new asset purchase program that has been popularly described as a second round of quantitative easing, or QE2. I thought it would be useful to take a look at what has actually changed during the first 3 months of QE2.

The essence of QE2 is that the Fed buys some longer-term Treasury debt and pays for it by creating reserves on which the Fed pays an overnight interest rate. In my view, in the current environment, interest-bearing reserves are for all practical purposes the same kind of security as a very short-term Treasury bill. The net effect of such a Fed operation is to lower the average maturity of the combined outstanding debt of the Federal Reserve and Treasury. One view of how such an operation might affect the economy is that a big enough drop in the net supply of long-term debt might result in a decline in long-term yields. In a research paper with Cynthia Wu, we found that historical changes in the maturity structure were associated with changes in the slope of the yield curve, and that those historical correlations were consistent with the claim that massive Fed operations of this type might have a modest ability to lower long-term interest rates in the current environment. A number of other researchers have also obtained similar results using a variety of methods. 

I noted in December that QE2 as actually implemented by the Fed differed from the particular scenarios that we and other researchers had looked at in two key respects. First, the Fed has primarily been buying debt of intermediate maturity (2-1/2 to 10 years) rather than the longest term debt outstanding. Second, the Fed spread these purchases out over a period of 8 months, during which time we could anticipate significant changes in the composition of debt issued by the Treasury which could potentially offset any effects of QE2.

The graph below provides our calculations of the average maturity of publicly-held debt both before and after the Fed’s operations, updated to include the first 3 months of QE2. The blue line is the average maturity (in weeks) of debt issued by the U.S. Treasury. The green line is the average maturity of publicly held debt, that is, the green line represents the results of subtracting off the Fed’s holdings of Treasury debt. Historically the green line was above the blue. This is because the Fed preferred to buy the shorter-term debt, as a result of which the average maturity of the remaining debt held by the public (green) was bigger than that for the debt as originally issued (blue). However, since the start of 2008, that relation has been reversed– the Fed has been buying a disproportionate share of the longer-maturity debt, and thus has been a factor in reducing the average maturity.


Blue: average maturity (in weeks) of marketable nominal U.S. Treasury debt outstanding as of the end of the month, 1990:M1-2011:M1. Green: average maturity of debt other than that held by the Federal Reserve.

But also since 2008, the Treasury has been issuing more long-term debt faster than the Fed has been buying it, so that the green line continues to rise over time. What we find in the latest data is that this trend has continued over the last 3 months, even with QE2. The graphs below highlight details of the last year. The top panel is the average maturity of publicly-held Treasury debt inclusive of all Fed operations, that is, it corresponds to the green line in the preceding figure. Although the average maturity in the second and third months of QE2 (December and January) fell a little below that for the first month (November), the average maturity in every one of these three months was bigger than in every month of the two years prior to QE2. The second panel shows the fraction of publicly-held Treasury debt (again, after netting out the Fed’s operations) that is of 10 years or longer maturity. This has gone on to make new highs in both December and January.


Top panel: average maturity of Treasury debt other than that held by the Federal Reserve, 2010:M1-2011:M1. Bottom panel: fraction of outstanding Treasury debt not held by the Federal Reserve that is of 10 years or longer maturity, 2010:M1-2011:M1.

Our conclusion is that if QE2 made a positive contribution to the improving economic indicators since the program began, it could not have been through the mechanism of shortening the maturity of publicly-held Treasury debt.

This does not rule out the possibility that QE2 had an effect through some other channels. Another possible mechanism is that, by announcing QE2, the Fed successfully communicated that it had a higher inflation target than some observers had assumed, and successfully communicated that the Fed had both the tools and the will to prevent outright deflation. It appears to be quite an accurate characterization that QE2 did have an effect on many people’s expectations. Indeed, some observers had quite a passionate response that I find hard to reconcile with the fundamentally modest nature of what the Fed has been doing. Using the tool of QE2 as a device for helping to manage expectations is in fact the main argument I can see for having the Fed rather than the Treasury be the agent responsible for announcing and carrying out the plan. Even so, I doubt that it can make much sense for the Treasury to pull so hard in one direction that it completely undoes any real effects of QE2.

But whether it makes sense or not, that’s what’s been happening so far.

Originally published at Econbrowser and reproduced here with permission.