In the past few months, our own measures of economic risk have remained persistently unfavorable, as have the indications from the Economic Cycle Research Institute (ECRI). It is worth emphasizing that we have no hope or interest in seeing an economic recession. Rather, that is a conclusion that the data - at least data that reliably associates with recession - forces on us. As I've previously noted, the typical lead time is in the range of 13-16 weeks. This isn't a hard-and-fast rule. Rather, you estimate the "lead-time" of any indicator by calculating the correlation between the indicator and subsequent data that is observed a varying number of weeks or months later, and then finding where that correlation reaches its peak. For our measures, the signals tend to lead the data by 13-16 weeks (3-4 months). It has been about 15 weeks since our main composites turned negative, so while every cycle is different, I suspect that we are on the cusp of observable economic deterioration.
There are undoubtedly lots of indicators that analysts can point at as anecdotal evidence that there is no recession risk. The issue for us is classification accuracy - does the data reliably discriminate between recessions and recoveries? A signal that says "it's always an expansion" will be correct most of the time. A signal that says "the state of the economy is always what it was last month" will always be right except at the crucial turning points, but relies on a "true" determination of recession or recovery that would not have been actually available in real time. What matters is the "confusion matrix" - how often was the observable signal correct when it turns out that there was an actual recession imminent or in progress, how often was the actual signal correct in periods that were later determined to include no recession, and how sparse are the mixed instances where the signal differed from the facts? For our part, we can't reconcile the data we are observing with an ongoing economic expansion. It's that simple.
To extend the evidence beyond our own measures and ECRI's analysis, the chart below presents data (through October) from the Organization for Economic Cooperation and Development, an international quasi-governmental agency that sets international standards on a wide range of economic policy issues. The OECD publishes its own set of leading economic indicators on developed and developing countries. Notably, we've never observed deterioration to the extent that we presently observe, except when the U.S. was in or entering a recession.
The chart below is somewhat noisier, but conveys what we observe internationally - a concerted global economic downturn is unfolding here - not simply one or two countries. I should note that the China figures from the OECD typically run much higher than other nations around the world, so they are plotted on the right scale. But even there, the growth rate has dropped to levels that are consistent with a concerted global economic downturn. The composite ("Total") signal has never been at present levels outside of recession. The uniform downturn in these indices (including China on an adjusted basis) is difficult to align with the concept that the U.S. economy faces a near-term expansion.
