Calculate risk Remind me that heavy-duty truck sales collapse during a recession.I would like to know how the 12 month change in this variable compares to the corresponding change in mileage (suggested Steven Kopitz). The latter fared pretty badly.
figure 1: Recession probabilities using 12-month period changes in heavy truck sales (blue) and vehicle miles traveled (tan). 50% threshold red dotted line. The NBER uses shades of grey to define the peak and trough dates of the recession.
The McFadden R2 for heavy-duty trucks returned 28 percent, while the vehicle mileage (VMT) was 7 percent. One can also easily see that using the 50% threshold, the truck regression captures all recessions after 1970 with no false positives, while the vehicle mileage regression will miss all recessions except 2020 unless a lower 25% threshold is used. Outrageous threshold (in this case, we have a false positive for 1995). A result of little change is the preferred measure using Kopits, the change in the 12-month moving average. Really, VMT is technically a bad coincidence indicator of recession.
In either case, there is no sign of a recession in May (13% vehicle miles traveled) or July (4% heavy trucks) (the latest figures for vehicle miles traveled and truck sales, respectively).
