I have had numerous comments on my previous article ECRI WLI: Slowdown or double-dip recession. They all centred on the following:
• When the ECRI WLI smoothed annualised growth rate reaches extreme levels a recession follows and equity prices fall.
• There is no relationship between the ECRI WLI smoothed annualised growth rate and the yield on U.S. long bonds.
Well, let me put my cards on the table. I believe we have found the algorithm for the smoothed annualised growth rate of ECRI. By applying the algorithm to the actual published WLI numbers since 1992 the result explained more than 99% of the actual smoothed annualised growth rate.
Sources: ECRI (various internet sources); Plexus Asset Management.
In my previous note I pointed out the relationship between the ECRI WLI and the S&P 500 Index and explained that the reason why the smoothed annualised growth rate of the ECRI WLI led the S&P 500 Index was the method of calculating the smoothed annualised growth rate. Well, here it is! The same algorithm applied to the ECRI WLI to calculate the smoothed annualised growth rate was applied to the S&P 500 Index. I repeat my question in my previous note: Is the historical relationship a coincidence?
Sources: ECRI (various internet sources); I-Net Bridge; Plexus Asset Management.
The relationship since the beginning of May 2003 has been particularly strong.
Sources: ECRI (various internet sources); I-Net-Bridge; Plexus Asset Management.
The reason why the stock market is so narrowly correlated with the ECRI WLI (R-squared = 0.8262) is that the stock market is a major component of the ECRI WLI, if not the most important component!
I have never disputed the fact that in the past the ECRI WLI smoothed annualised growth rate led to recessions when the rate dropped to extreme low levels. What I warned about is that if the stock market bounces and goes higher in the coming weeks the ECRI WLI smoothed annualised growth rate will bottom and gain traction. What is more is the fact that if the stock market continues further south the market’s smoothed annualised growth rate will decline further, taking the ECRI WLI smoothed annualised growth rate down with it. It is the stock market making the calls on a coming recession and not the ECRI WLI!
When the same algorithm is applied to calculate the smoothed annualised growth rate of the yield on the 10-year government bond it is clear that the yield has had a bearing on the WCRI WLI smoothed annualised growth rate, especially since May 1998.
Sources: ECRI (various internet sources); I-Net-Bridge; Plexus Asset Management.
When the smoothed annualised growth rate of continuing jobless claims is compared to the ECRI WLI Smoothed Annualised Growth rate it is evident that the former is another major input factor in the calculation of the latter. (Please note that the continuing claims are in reverse order).
Sources: ECRI (various internet sources); FRED; Plexus Asset Management.
Shorter term:
Sources: ECRI (various internet sources); FRED; Plexus Asset Management.
Other factors such as money-supply growth also influence the ECRI WLI smoothed annualised growth rate. The actual components and weights in ECRI’s “black box” are unknown but it is clear the stock market, bond yields and jobs are probably the factors to watch − not the actual numbers, though, but the smoothed annualised growth trend.
Finally, here is a gauge of this week’s ECRI WLI smoothed annualised growth rate (to be released on August 6, 2010) based on data for the the week ended 30 July 2010:
