Thursday, December 2, 2010

Reply to amv on Sumner (fg)

1. You are correct, I implicitly assume a zero NGDP growth rate target becasue this is what I can read from Sumner's latest post; albeit I know that Sumner wants to target a growth rate. However, his benchmark is the level of NGDP in early 2008, which he compares with the level in 2010Q3 (and not with the implied path, at least he does not refer to it) in order to draw a conclusion on the policy stance. I just made the same analysis with Germany instead of the euro area (btw: the officially targeted price index is HICP and not the core price index).

2. The more interesting question to me is to ask why he choses the 2008Q1 date as the benchmark/initial value? When you look at this figure, you may find that NGDP has been growing app. at a rate between 2.5 and 3% from 1997Q1 until 2005Q4; afterwards, we may call the period from 2006-2008 as a (trend-) break which is reflected by the hump-shaped pattern of the time series. What is, if the the ECB/Buba targets a NGDP growth rate of 2.5-3% (ex post)? During 2006-08, the growth rate has been above 5% and all indicators pointed toward rising inflation expectations. In 2010Q3, NGDP is back on track, if you project the implied 2.5-3% target path when taking 1997/98 as initial value (provided that you do not allow for base drifts).

UPDATE: The same observation holds if you take the euro area record 1999-2005 as benchmark with an average (and stable) NGDP growth rate of 3,96. The period 2006-2008 is characterized by growth rates of around 5.3%. If we project the implied NGDP target rate from 2005 onwards, we are on track without base drift.