The change in ECRI’s Weekly Leading Index is now a hair’s width from their recession guarantee. The growth reading moved from -8.3 on the July 2 reading to -9.8 on the July 9 reading (released today).
I have to say that -10 is pretty arbitrary and an example of a whole number bias, as -9.8 has actually been a more reliable indicator. I just went through their data, and April of 2001 (which was later backdated by NBER to be in recession) never quite hit -10, but did cross -9.8. So -10 has may have no false positives, but -9.8 also has no false positives and one fewer false negative. Since their calculation of this number is mostly voodoo to me, I won’t weigh in too heavily, but statistics are statistics, and this is a more reliable indicator than ECRI’s own “guarantee.”
