Following Friday’s mediocre jobs report, the Atlanta Fed GDPNow Forecast made a surprising leap.
The GDPNow model forecast for real GDP growth (seasonally adjusted annual rate) in the first quarter of 2016 is 2.2 percent on February 5, up from 1.2 percent on February 1. After this morning’s employment report from the U.S. Bureau of Labor Statistics, the forecast for real consumer spending growth increased from 2.5 percent to 3.0 percent and the forecast for real gross private domestic investment growth increased from -0.4 percent to 2.1 percent.
I pinged Patrick Higgins, senior economist at the Atlanta Fed and he provided the following explanation:
About 30 series from the employment situation report are used to calculate the dynamic factor that feeds into the forecasts of the monthly source data for the model. The January 2016 value of the dynamic factor was -0.73 after the February 1 GDPNow update and is 0.04 as of today. The factor has mean 0 and standard deviation 1, so the current estimate for January is very close to average. To provide some historical context, the value of -0.73 was about in-line with values seen in late 1991 and early 2003.
It is technically complicated to decompose how much particular variables in the employment report are responsible for the rise in the dynamic factor. One thing that can be said, however, is that it’s not due to average hourly earnings growth since that variable is not in the model. Also less weight is being put on the January ISM Manufacturing report since we now have more January data than we did on February 1.
It’s not easy to predict which way the Nowcast will move following economic data releases. I expected lower because of bad economic reports across the board. What seems to have happened is the initial forecast was simply too pessimistic.
This does not change my assessment of the economy. GDP revisions are the norm, especially at turning points, and it seems likely the economy is at that turning point.
Mike “Mish” Shedlock