by Calculated Risk on 9/15/2014 11:35:00 AM

Excerpts from a note by Goldman Sach's chief economist Jan Hatzius: Q&A on "Why Renege Now?"

Q: You expect the phrase “considerable time after the asset purchase program ends” to remain in the statement. Many others don’t; what are they missing?

A: Many are missing the distinction between a decision not to extend existing guidance and a decision to renege on existing guidance. Let’s compare the current situation with the runup to the last rate hike cycle, when the committee went from “considerable period” in August-December 2003 to “patient” in January 2004, “measured pace” in May 2004, and finally rate hikes in June 2004.

The shift from “considerable period” in December 2003 to “patient” in January 2004 is an example of a decision not to extend existing guidance. Informed observers concluded from this shift that the committee had retained its guidance that a hike would probably come no earlier than June, but was unwilling to go beyond that. And indeed, the first hike came in June.

But if the committee removed the phrase “considerable time after the asset purchase program ends” this week or replaced it with something weaker, it would not only decline to extend the existing guidance into the future, but would in fact renege on the existing guidance. That would be a much bigger step than in January 2004.

Q: But don't they have to change the guidance as QE ends?

A: Eventually yes, but we think September is too early because QE has another six weeks to run, assuming they taper the program down to $15bn per month this week and end it at the October 29-30 meeting.

Even a material change at the October meeting would be a shortening of the existing guidance. (By "material change" we mean anything that goes beyond deleting the phrase "after the asset purchase program ends.") For example, replacing "considerable time" by “some time” on October 30 might be interpreted to mean that the no-hikes guidance expires, say, in February instead of April. Of course, it is possible that the data surprise sufficiently on the upside or the outlook changes in some other way to justify a more material change at the October meeting--remember, the guidance is conditional on the outlook.

But barring such a surprise, the right time to make a substantive change in the guidance is the December meeting. There are three basic options at that point: 1) keep “considerable time” and effectively extend the no-hikes guidance past the end of April, 2) move to weaker terms such as “patient” or “some time” and thereby decline to extend the no-hikes guidance past the end of April, or 3) move to more qualitative guidance phrased in terms of the remaining amount of slack or the level of inflation relative to the target.
...
Q: Do you think your forecast implies a dovish surprise for the markets this week?

A: Probably, but not necessarily. It is clear that many market participants expect a change in the "considerable time" language. If we are right that the language will remain unchanged, this would likely be dovish for near-dated fed funds futures contracts. That said, there are a lot of moving parts in an FOMC meeting that includes a press conference and a new SEP. If other aspects of the statement such as the post-liftoff guidance sound more hawkish, that could negate or even overwhelm the impact of the liftoff guidance, at least for the longer-dated contracts. Also, regardless of any changes in the statement, we expect the "dots" to drift up a bit further in 2015-2016 and to show rates for 2017 that are well above current market pricing, although this already seems to be widely expected and might therefore not have much impact on its own. And finally, of course, we might just be wrong and “considerable time” might go after all.