Recession Predictor & A Warning Signal!

[J. Martin]

As I was browsing through data for potential recession predictors, I ran across none other than an index specifically designed to determine when recessions are taking place! The index is actually designed to estimate if we are CURRENTLY in a recession, but every time it registers more than 4% for two consecutive months, we end up already in a recession or go into one within 6 months (over the last 48 years and 7 recessions). It just registered over 4% for the first time since the last crisis in December 2015!

What was striking to me about this data is that EVERY time the reading is over 4.0% for two consecutive months, a recession has either already started, or starts within 6 months. EVERY TIME for the last 48 years and 7 recessions. It has never registered over 4.0% for two consecutive months and NOT had a recession. We had our first in December 2015! So I will be keeping a close eye on the Jan 2015 number, and sent an email to the publishers of the data to pick their brain on my analysis..

Note that this is not the type of index that will slowly rise as things get worse. It rises quickly when recessions are starting, and we should know within 1-3 months if we are very likely progressing into a recession. The index uses non-farm payroll employment, the index of industrial production, real personal income excluding transfer payments, and real manufacturing and trade sales - along with some fancy statistical transformations.

Let's look at it recession by recession..
T-0 is the month of the second consecutive occurrence of the index over 4.0% "D Day"
T-1 is the month of the first occurrence.
Orange is the start of the recession.

This predictor is very close to the actual start of recessions (even though recessions starts are not announced by the NBER until significantly later.) This makes sense, because that is the objective of the index - to identify recessions earlier than the NBER method. The table below shows the number of months until the start of the recession, as measured from the second consecutive monthly reading above 4.0%. (Note that 2007 is one month off..)

Would it concern you if we have another reading above 4.0% in Jan 2016?
Is this worthless?
Some informational or predictive value?