Jan Hatzius Q&A On QE2
Q&A on QE2 (Hatzius), (via Tungstenman Sachs)
Our view remains that the Federal Open Market Committee (FOMC) will once again ease monetary policy via unconventional measures in late 2010 or early 2011. Our views have not changed, and today’s comment discusses them in Q&A form. We believe that purchases of US Treasury securities cumulating to $1 trillion or more are the most likely cornerstone of the program; that the September 21 FOMC meeting is probably too early for a big announcement, but that November 2-3 is a possibility; and that it would likely “work” to a limited degree, perhaps boosting real GDP growth by a little under ½ percentage point per $1 trillion in purchases.
Q: Have you changed your expectations for Fed policy in recent days?
A: No. In early August, we adopted a view that the Federal Open Market Committee (FOMC) would ease monetary policy further via “unconventional” measures. The most likely option, in our view, was a return to a large-scale asset purchase program focused on longer-term US Treasury securities cumulating to at least $1 trillion (see “Climbing Aboard QE2 to Avert a Double Dip?” US Economics Analyst, 10/31, August 6, 2010). At the time, we indicated that the most likely timing for this shift would be late 2010 or early 2011. This remains our expectation.
The rationale for this forecast is that we expect the FOMC’s output and employment forecasts to converge to our own. (This is the conceit inherent in any monetary policy forecast.) Thus, we believe they will move to a view that real GDP will only grow at about a 1½% (annualized) rate through early 2011 and the unemployment rate will rise to 10% by the spring of 2011. If so, they would view growth as clearly below trend according to both the GDP and employment data, and the output gap as widening anew. Given the recession risk that a rising unemployment rate has reliably indicated in the past, we believe that the committee will respond by easing policy further.
Q: What specific measures do you expect?
A: There are two main options, although they are by no means mutually exclusive: (1) large-scale asset purchases and (2) changes in Fed communications. Fed officials believe that these would ease financial conditions via, respectively, a lower “term premium” at the long end of the yield curve and a lower path for the expected level of short-term interest rates. Other options include (3) a lower interest rate on excess reserves, (4) a higher and perhaps more explicit Fed inflation target than the current 2%, (5) setting explicit ceilings on longer-term Treasury yields, to be enforced by promising to buy unlimited amounts, and (6) lending to nonbanks under Section 13 (3) of the Federal Reserve Act, presumably with funding provided by Congress. Chairman Bernanke gave relatively short shrift to measures (3) and (4) in his speech at the 2010 Jackson Hole Symposium, and we likewise regard measures (5) and (6) as unlikely in the absence of another clear recession and/or a renewed financial crisis.
Between options (1) and (2), the FOMC appears to favor renewed asset purchases. We believe that the main reason is that it is difficult to think of a plausible way to strengthen of the commitment beyond the current phrase that “…economic conditions…are likely to warrant exceptionally low levels of the federal funds rate for an extended period.” The next step in terms of the statement itself would be a commitment—at least a conditional one—that the federal funds rate will not rise until a certain date (e.g. mid-2012) or until a certain economic conditions has been met (e.g. core PCE inflation back above 2%). The FOMC is likely to be reluctant to provide such an assurance, although less formal speeches and testimony explaining why the funds rate may need to stay low for a “really extended” period could be used to buttress the impact of any QE2 announcement.
If the primary measure is a return to asset purchases, the question becomes what type of assets Fed officials will decide to buy. We believe that US Treasury securities are the most likely option. This is partly because the Fed already owns a relatively large proportion of the agency MBS market, and partly because many Fed officials (not just the “hawks”) have concerns about interfering with the allocation of credit via buying anything other than Treasury securities. Treasuries have the additional advantage of keeping the Fed’s portfolio as liquid as possible, facilitating the ultimate return to a more normal-sized balance sheet.
Finally, as we discussed in a daily comment last week, there are three potential approaches to reinstituting a Treasury purchase program: (1) a “big bang” announcement of a large number up front, (2) a small announcement with a clear indication that the committee will buy more if economic conditions warrant it, and (3) a fixed monthly or quarterly amount of purchases that continues until certain economic conditions are met. (See Jari Stehn, “Big Bang versus Small Steps: Thoughts on Designing QE2,” US Daily Comment, September 8, 2010.)
Q: Can you be more specific about the timetable for the big QE2 announcement? When will it happen?
A: There are five FOMC meetings between now and the end of the first quarter. A significant announcement at next week’s meeting seems unlikely to us. Given the starting point for the FOMC’s economic forecast—more optimistic on growth and higher on inflation than our own—and the somewhat better-than-expected data over the past few weeks, we do not believe that either of the two criteria spelled out by Bernanke at Jackson Hole—a “deviation from price stability in the downward direction” or a “significant weakening of the economic outlook”—has been met yet.
However, a move could occur at any of the subsequent meetings, but it will very much depend on the data and the implications of the data on the Fed’s forecasts. In particular, a move at the November 2-3 meeting is a possibility if the data warrant it, despite the fact that the FOMC statement will be released less than 24 hours after the polls close for the midterm congressional elections. Historically, there is not much evidence that midterm elections have much impact on the timing of monetary policy announcements, especially on the side of easier policy. For example, the FOMC cut the funds rate by 50 basis points on November 6, 2002, a day after the midterm elections.
Q: Will there be many dissents against a return to unconventional policies?
A: At this point, we don’t think so, especially if the FOMC decides to purchase Treasury securities. We suspect that Kansas City Fed President Hoenig would dissent yet again, but that may well be all. Although a recent article in the Wall Street Journal noted some discomfort with asset purchases even among Washington-based Governors Duke and Warsh, we believe that they would ultimately accept the chairman’s leadership on such an important decision. But of course, all this partly depends on how clear-cut the decision will look in terms of the incoming economic data, and how “proactive” the Fed leadership will want to be in erring on the side of earlier rather than later action. One cost of being more proactive is that it may entail more disagreement.
Q: Will it work?
A: Kind of. Our analysis of the first round of asset purchases announced in late 2008 and early 2009 found that they pushed down 10-year Treasury yields by about 25bp and eased our Goldman Sachs Financial Conditions Index by 80bp per $1 trillion (see Jari Stehn, “Unconventional Fed Policies and Financial Conditions: How Tight a Link?” US Daily Comment, August 17, 2010). In a second round, we believe that the impact on financial conditions would be smaller, mainly because the credit spreads are much tighter than they were back then. If the impact on financial conditions is 50-60bp rather than 80bp, this would imply a potential boost to real GDP growth of a bit under ½ percentage point, using historical rules of thumb. That’s big enough to matter for economic forecasters, but it is hardly enormous.