By Dave Altig
Lots of bloggers have expressed lots of different views on what they consider interesting in Roger Lowenstein's new profile of Federal Reserve Chairman Ben Bernanke in The Atlantic. Authors at The Big Picture, Free Exchange, Marginal Revolution, Modeled Behavior, and Real Time Economics, for example, all highlight distinct passages that piqued their individual interests. This is what caught my eye:
"Bernanke's unconventional programs have been implemented in two phases. During the financial crisis of 2007–09, he bailed out a handful of large banks and devised a series of innovative lending operations to disperse credit to banks, small businesses, and consumers (virtually all of these loans have been repaid at a profit to taxpayers). He also lowered short-term interest rates to nearly zero and made private banks run a gantlet of stress tests to ensure some minimal level of solvency going forward. Although fierce anger against the bailouts persists, there is little argument that this first stage was a success. However untidily the rescue was managed, the financial crisis is over.
"In the second stage, Bernanke has sought to revive a weak economy by maintaining short-term interest rates at close to zero, and by purchasing, in vast quantities, long-term Treasury bonds and mortgage-backed securities. This second phase has been, if anything, more controversial than the first. Its success is much harder to measure (we have no way of knowing whether the economy's improvement would have been less robust, and how much so, without Bernanke's efforts). And it has exposed Bernanke to charges of meddling too deeply in the private sector, of disrupting the economy's natural rhythms long past the point when such intervention is necessary. In particular, critics note that the Fed has stuffed the banking system with $1.5 trillion in excess reserves—money for which the banks have no present use, loan demand being modest, but which could one day spark an epidemic of inflation.
"Michael Bordo, a monetary historian at Rutgers, told me that in this second phase, 'Bernanke has moved into areas that were quite different from what the framers had in mind. One of the risks the Fed is facing is of overreach.' "
I want to point out up front that I respect Bordo a great deal. He is a friend and collaborator to many of us at the Atlanta Fed. In fact, he was the co-organizer of our 2010 conference that commemorated the 1910 Jekyll Island meeting that resulted in draft legislation (the Aldrich Plan) for the creation of a U.S. central bank. He is also the co-editor of the proceedings volume from that conference that will be forthcoming from the Cambridge University Press. But I may disagree with his assessment here.
First, just to dispose of the easy stuff, there probably isn't much about "normal" monetary policy that the framers of the Fed had in mind. Legislation creating the Federal Open Market Committee (FOMC) didn't arrive until 1935, and interest rate policy aimed at addressing broad macroeconomic conditions was not likely of much concern to folks preoccupied with birthing a lender of last resort.
But I don't think that is what Professor Bordo has in mind. I think the criticism is that the "purchasing, in vast quantities, long-term Treasury bonds and mortgage-backed securities" represents a discrete break from normal, well-established, plain-vanilla monetary policy as it was understood pre-crisis.
There are generally two aspects to this criticism, one related to the type of assets that the Fed has purchased and one related to the sheer size of those purchases. The first of these criticisms amounts to the contention that the FOMC should restrict itself to dealing in Treasury securities alone and that the foray into the mortgage-backed security market represents an unwise exercise in credit allocation. Personally, I have some sympathy with this concern (articulated over a decade ago by former Richmond Fed president Al Broaddus and former Richmond Fed research director Marvin Goodfriend), but I have also expressed my view that such forays might be understood as substitutes for normal policy in conditions in which normal policy is constrained by circumstances.
And so it is with the scale of the Fed's asset purchases. Normal policy in normal times would drive asset purchases in the service of generating desired movements in short-term interest rates (leading, it is assumed, to effects on a broader set of asset yields). This particular routine is obviously not available when the federal funds rate is at its zero lower bound. But that does not mean the logic of this mechanism is absent in asset purchase programs implemented once that bound is hit. Specifically, one way to think about large-scale asset purchases is that those purchases can replicate the effects of federal funds rate reductions, if only those rate reductions could be implemented.
There are several ways to get at this question—one very nice summary has been provided by Sharon Kozicki, Eric Santor, and Lena Suchanek from the Bank of Canada. At the Atlanta Fed, we have our own version of gauging the effects of large-scale asset purchases. Our approach involves estimating a "virtual federal funds" based on the size of Federal Reserve asset holdings relative to total commercial bank liabilities. The details of how this virtual funds rate is constructed can be found here, but the upshot is in this chart:
Two points. First, the virtual funds rate, which combines the zero actual funds rate with the estimated interest-rate equivalent from the Fed's cumulative asset purchases, is about 200 basis points to the south of zero. Second, as a reference point, the chart compares our virtual funds rate to one version of the Taylor rule that relates the federal funds rate to deviations of gross domestic product from its estimated potential and inflation from the Fed's 2 percent long-run objective. As the chart depicts, this comparison is, as of now, just about right, despite the fact the federal funds rate is constrained near zero.
I hasten to add that the Taylor rule in the chart above is, for present discussion, for reference only. There are certainly plenty of arguments as to why the rule depicted in the chart may not provide the right policy guidance. For example, there are debates over how to appropriately specify the Taylor rule, as reflected in this post by David Papell at Econbrowser. Furthermore, there are arguments as to whether policy more generally ought to be more aggressive at present than any standard Taylor rule prescription.
My intent is not to argue whether policy is, at present, appropriately calibrated. My point is that rather than thinking of large-scale asset purchases as unconventional policy, perhaps we should think of them as conventional policy in unconventional circumstances.