January 10, 2012
Eliminating the false positives is not without cost – the previous method using WLI growth metric gives you more warning of the 2nd, 3rd, 4th and 7th recessions. You will want to use both methods when evaluating risks of recession. If the first method, using the WLI growth computed by ECRI, is flagging recession, then you can treat it is a valid warning but subject to false positives. You may elect to act on this warning or revert to the second method for further confirmation (sacrificing a few weeks in the process if a recession is indeed iminent).
As far as other monthly leading indexes go, as of January 2, 2012 the e-forecasting.com eLEI fell sharply out of its recession zone. It has much better recession dating performance metrics since 1967 than either WLI model discussed above, as we can see in the below graph. (Note, however, that the single false positive shown in the performance metric box below is on top of the one eLEI seems to have just made.)
The other widely followed US LEI is the Conference Board, which has flatly refused to recognize any heightened risks of recession recently, sitting at almost zero probability of recession. This LEI has similar NBER recession forecasting performance metrics to the two ECRI WLI models discussed above. In fact, it is also prone to occasional false positives and yet is showing a reading decidedly more optimistic than either WLI model.
The Conference Board has announced extensive benchmark revisions to its LEI, which are supposed to improve its ability to detect economic cycle turning points. This is the first major overhaul of the components of its LEI since 1996, when The Conference Board assumed responsibility for the business cycle indicators program from the Bureau of Economic Analysis at the U.S. Department of Commerce. The new time series is scheduled for release on January 27, and we will perform a separate analysis on the performance of the new LEI with respect to old LEI as well as the WLI and eLEI when that new information is released. We may yet announce a new king of the LEI’s in that analysis!
To put the power of using multiple indicators into perspective, my firm’s composite co-incident SuperIndex, which is made up of nine other metrics, including the WLI, and which I covered in a prior article, has an AUC accuracy of 96.7%, an NBER Capture rate of 100% with zero false positives since 1967. Also, a simple “Diffusion Index” which tracks how many of the 9 SuperIndex components are in their respective recession territories has an AUC of 97.3%, an NBER capture of 98.8% and zero false positives since 1967. Both these methods have extremely powerful recession dating prowess and although they appear similar performance-wise there are however nuances with lead and lags into and out of recession. We show the performance of the Diffusion Index below for the last two recessions – as you can see the index fell from 2 to 1 last week when the eLEI fell (rather sharply) out of recession territory, leaving just the WLI calling recession. The model calls recession when more than four indicators are in recession territory and then calls the end of recession when the number of indicators in recession falls below 3. Given the historical accuracy of the diffusion index it is extremely hard to take the current ECRI WLI recession signal seriously in the near term.
For PART-2 of this article “Further improving the use of the ECRI WLI”, click here.
Dwaine van Vuuren is CEO of PowerStocks Investment Research, a South African-based provider of investment research and specialist in dating US recessions If you would like to receive the next 4 weeks SuperIndex Recession Reports for free, just email us at with FREE SUPERINDEX in the subject line.
Display article as PDF for printing.
Would you like to send this article to a friend?
Remember, if you have a question or comment, send it to .
