As Zero Hedge said back in January, when we predicted the transitory "bounce" or "temporary hard spot" in the economy, well ahead of everyone, the first thing that would need to happen for QE3 to be launched is for Goldman's Jan Hatzius to admit that his December 2010 call for an American golden age was a disaster, and to recognize that the economy is now contracting at a rate that will put last year's Q2 and Q3 GDP drops to shame. As of two weeks ago, that has happened. The only thing that was missing from our checklist was for Hatzius replacement, who enjoys the occasional Hefeweizen at the Pound and Pence with his former co-worker Bill Dudley, to make it clear what Goldman's position vis-a-vis QE3 is. Well, as of a few minutes ago we can cross that box too after Hatzius' lieutenant Sven Jari Stehn just said that "A sharp increase in the Fed's assessment of recession risk would most likely trigger significant additional monetary easing even if inflation remains well above their target." What has inspired this change in hear? Simple: nothing less than the "realistic–possibility of a significant further deterioration in the economic outlook." Oh well, the recovery was fun while it lasted. Same for the strong dollar. Next up: Jackson Hole in one month, where Bernanke will announce what the Dudley-Hatzius tag team has been whispering in his ear for the past month.
From Goldman: What Do Taylor Rules Say About the Threshold for Monetary Easing? (Stehn)
Discussion about additional monetary easing has reemerged with weak growth in the first half of the year. Fed Chairman Bernanke, for example, said that persistent weakness in activity and renewed deflationary risks would “imply a need for additional policy support” during his Congressional testimony on July 13. More recently, Chicago Fed President Charles Evans said that he might support a third round of asset purchases.
What do Taylor-type rules say about the need for additional monetary easing? Our baseline Taylor rule describes how Fed officials have historically set the funds rate using four-quarter-ahead forecasts of core PCE inflation and the unemployment gap (actual less “structural” unemployment). (For details see Jan Hatzius and Sven Jari Stehn "The Warranted Funds Rate: Is It Really Negative?" US Daily, March 10, 2010.) Under our forecasts this rule currently suggests that the "warranted" funds rate is still substantially negative at -3% (see Exhibit). (Under the Federal Open Market Committee's (FOMC) June forecasts--which are somewhat stale at this point--the warranted Funds rate is also negative but higher at -1.4%.) This prediction is close to that from a rule estimated by Glenn Rudebusch of the San Francisco Fed, an update of which suggests that the funds rate "should" currently be around -3.3% (see "The Fed's Exit Strategy for Monetary Policy," FRBSF Economic Letter, 2010-18).
The problem with these simple rules, of course, is that they do not take into account the Fed's unconventional policies. By lowering long-term yields, the Fed's asset purchases were designed to substitute for conventional monetary policy and thus imply that the appropriate funds rate should be higher than suggested by the simple Taylor rules discussed above. (Put another way, if we "give the Fed credit" for asset purchases, the funds rate itself does not need to be as low.) Similarly, the Fed's guidance that it would keep interest rates unchanged for an "extended period" was designed to ease financial conditions. Our analysis suggests that the Fed's asset purchases are equivalent to a cut in the funds rate of about 75 basis points (bp) per trillion (trn) dollars of purchases, while the "extended period" language is worth an additional 30bp. (For details see Jan Hatzius and Sven Jari Stehn,"QE2: How Much is Needed?" US Economics Analyst, October 22, 2010.) Adjusting our baseline Taylor rule for these estimates suggests that the appropriate funds rate is currently -1.1% and will turn positive in mid-2013 (see Exhibit).
Fed Chairman Bernanke's statement during the Congressional testimony suggests that Fed officials see asset purchases as a more powerful substitute for conventional monetary policy than our estimates suggest. Specifically, he argues that the Fed's asset purchases are equivalent to a 67 to 200bp reduction in the funds rate per $1trn of purchases. These magnitudes are based on a number of studies that suggest that the second round of asset purchases lowered longer-term interest rates by approximately 10-30bp. Unfortunately, Bernanke does not explain how he translates the estimated long-term reduction in interest rates into funds-rate equivalent terms. We suspect that these estimates are based on a study by Glenn Rudebusch (see citation above), who argues that long-term interest rates are four times as powerful in affecting output as changes in short-term interest rates--exactly the same multiple that Bernanke uses. Adjusting our simple Taylor rule with the mid-point of Bernanke's range suggests there is no need for additional easing: the current monetary stance is appropriate and the funds rate should actually be raised soon (see Exhibit).
Under what circumstances would these rules call for significant additional monetary easing? One possibility is a much smaller estimate of the effectiveness of asset purchases than Chairman Bernanke's. Indeed, our own estimates imply that the impact of QE2 is only slightly more than half the midpoint of Bernanke’s range. Similarly, updated simulations of the Fed's macro model ("FRB/US") suggest that reductions in long-term interest rates have an effect on output that is 2½ times larger than equivalent declines of the federal funds rate after four quarters. These simulations suggest that the observed decline in long-term rates is equivalent to a 25-75bp reduction in the funds rate--also quite a bit lower than the magnitudes cited by Bernanke. But even with these estimates, the Taylor rule would not point to the need for significant further easing.
The second–and probably more realistic–possibility is a significant further deterioration in the economic outlook. We estimated a few weeks ago that it would take a 1¼-point increase in the expected unemployment rate or a 1-point drop in the inflation forecast for additional easing to become appropriate. (See Sven Jari Stehn, "Sizing the Fed’s “Zone of Inaction",” US Economics Analyst, June 17, 2011.) Moreover, it is possible that these rules overstate the hurdle against additional easing. After all, Fed officials are well aware that the US business cycle is highly asymmetric. In the postwar data, every increase in the unemployment rate by at least one-third of one percentage point (on a 3-month moving average basis) has indicated a recession. Therefore, the Fed would likely respond asymmetrically to improving and deteriorating information at this point. A sharp increase in the Fed's assessment of recession risk would most likely trigger significant additional monetary easing even if inflation remains well above their target.