Enough of all of these hand waving, finger to the wind predictions. (nothing sillier than an economist dreaming up 40% chance of recession, or even worse, a think tank or politician claiming 15% chance of some war happening)
Here is a forecast of an actual number, backed by more actual numbers, of tomorrow’s month-over-month % change in Personal Consumption Expenditures for April (PCE, released May 30) using April’s Advance Retail Sales, also MoM % change (ARS, released May 15). Data from FRED, both seasonally adjusted.
PCE is a measure of consumer spending (in real dollars), a huge part of the US economy.
As you can see from the below time series, and the scatter plot of PCE vs ARS, the two metrics are very correlated.
ARS for April was about -0.12% MoM. Using the ol’ eyeball regression approach on the scatterplot (imagine a straight line through the middle of the scatter, then look for the value of -0.12% on the x axis, go up vertically until in middle of the cloud of points where your imaginary, but realistic, line is, then move horizontally left to the y axis and read off a prediction for PCE for April).
(If you are enthusiastic, you can crunch the numbers exactly and produce a number to a few more decimal places.)
I get a forecast of about +0.4% PCE MoM for April (again, real dollars, seasonally adjusted). There is a cloud of PCE values near ARS = -0.12%, so obviously there will be uncertainty associated with this estimate, say a range of PCE MoM % of about 0.2% to +0.6%.
Edit: Above (and below graphs, etc) PCE is nominal, not real (not inflation-adjusted). There is also a real PCE figure that is released, which is inflation adjusted.
Let’s check tomorrow’s release from BEA to compare the above forecast versus actual.
