EJ Antoni (Heritage) is dubious about GDPNow’s (and other nowcasts) regarding Q3 growth. From X aka Twitter today:
Latest Q3 Nowcasts: ATL 2.5% (2.0% prev) NY 2.49% (1.94 prev) STL 2.05% (1.65 prev) It’s going to take stellar consumer spending numbers for Aug and Sep to make this a reality:
I wondered about this, in light of the measured acceleration of real consumption growth (m/m AR 4.6% in July). GDPNow reports nowcasted growth of components, like consumption so one can check the nowcast. GDPNow’s 30 August has a 3.8% q/q AR growth rate penciled in. What’s the picture look like?
Figure 1: Consumption (bold black), GDPNow nowcast as of 30 Aug (blue line), August SPF median (red line), and ARIMA(0,1,0) forecast (blue line), +/- 1 standard error band (gray lines), all in bn.Ch.2017$, SAAR, log scale. Forecast estimated in log first differences, 2021M07-2024M07. Source: BEA, Philadelphia Fed, Atlanta Fed, and author’s calculations.
Admittedly, GDPNow’s nowcasted consumption is above the August SPF median (although this forecast was generated in early August). However, based entirely on a simple autoregressive model, 3.8% q/q AR consumption growth would not be surprising at all, statistically speaking.
