The revisions of real GDP growth for 2015 I was announced this morning: down to –0.7% from 0.2%.
That’s in recession territory, but is not yet a recession. It’s not that unusual for the economy to dip that low, or for downward revisions to be that big. To be a recession, those low numbers have to be sustained.
Enter GDPNow from the Atlanta Fed. This is their new-ish real time forecast of the economy. It showed a fairly severe drop-off starting around the beginning of February.
But, there’s a trick to interpreting their data announcements: they keep forecasting last quarter’s real GDP growth until the initial announcement comes out about a month after the end of the quarter. So if you go look at their data, all through January they are forecasting 2014 IV. It’s only with the start of February that they start to forecast 2015 I, whose initial draft won’t be available for 3 more months.
So, in saying that their numbers start to look bad around the start of February, they are saying that 2015 I looked bad from the get go.
And GDPNow continues to show that weakness all through their forecasts for 2015 II that they started about 4 weeks ago.
First, a note about these charts. What I’m showing is the cumulative average forecast for each quarter, from the beginning of forecasts for that quarter. What that means is that the big vertical jumps, like the in the blue line in the top center, just mean that one quarter ended and another began. The jump is then the difference from the last forecast for the previous quarter to the first forecast for the next quarter.
I’m also showing moving averages of all the forecasts for a quarter. This means that starting from a jump, as the jagged line moves to the right it becomes more accurate, before there’s another jump to the next quarter.
This is what really worries me. This shows the components of GDP and their contributions to the forecasted growth rate of real GDP. Not surprisingly, consumption (PCE) is the largest, imports are negative, and so on.
What happened in 2015 I is that investment, government spending, and exports are all weak, and consumption is merely average.
Note that the above is just the forecast from GDPNow. The revision that was announced this morning is that the sum of these 5 components. GDPNow was a weak 0.9 when forecasting stopped at the end of April, and the revision is a 1.6 below that.
Why was it that weak? Exports were revised to –1.0 (1.1 below forecast), and PCE was revised to 1.2 (0.4 below the forecast). Basically, everything was weak to begin with, and the two blue components came in much lower than forecast. Both of those components are purchases of goods made in the U.S., one by foreigners and one by residents.
This chart shows the deviation from the mean for that component. And this really scares me.
Note the huge drop in the investment forecast. This is for the current quarter, not the one whose downward revision was just announced. This suggests that businesses have gotten the message that neither residents or foreigners are buying their goods, and so they’ve scaled back their investment spending severely.
This does not look good.
I tend to be an optimist. I make a point of remarking in my macro classes that I tend to miss business cycle peaks because of that optimism. I wonder now if everyone else is missing this one. We’ll see in a few months I guess.