Earlier I took a quick look at the most popular financial news websites, which includes Business Insider (certainly the most entertaining of the bunch), and I was struck by this headline: DAVID ROSENBERG: Here's Your Big Red Flag That We Could Be Heading For Recession.
I always find Rosenberg's chronically bearish commentaries of interest and in this case by the fact that he's reported to view CAPEX as a recession indicator. The Business Insider website included a chart illustrating the year-over year change in the 3-month moving average back to early 2003. Thus the chart only includes one NBER-designated recession on which to evaluate the efficacy of CAPEX as a recession indicator.
But I was also arrested by the fact that after the latest durable goods report was released, I had also charted the YoY change the 3-month moving average of CAPEX. It was an interesting chart, I thought, but inconclusive. First, let's take a look at monthly CAPEX, the popular abbreviation for Capital Expenditures, which the FRED database labels as Manufacturers' New Orders: Nondefense Capital Goods Excluding Aircraft. The data only goes back to February 1992, but in charting the complete series, my version adds another recession, the eight-month contraction in 2001, to the CAPEX context.
(click to enlarge)
Here is a year-over-year percent change of the series. (click to enlarge)
And, finally, the chart below is the YoY of a 3-month moving average of the complete series. (click to enlarge)
Indeed, the CAPEX monthly data has been negative for six of the last 12 months, and the 3-month MA has been trending down since March of this year.
Ultimately my sense is that this data series manipulation (YoY of the 3-month MA) has an insufficient track record to be considered a definitive recession indicator. N=2 is not enough to make reliable recession probability forecasts. But CAPEX is definitely something I'll add to my monthly Durable Goods updates.
The next Durable Goods update is scheduled for October 25th.