Home-Made Payrolls

This post appeared yesterday on RealMoney. Click here for a free trial, and enjoy incisive commentary all day, every day. As the third-quarter earnings season winds down, the primary catalyst for U.S. equity markets over the remainder of the year will likely be the tepid recovery in the U.S. labor market, reason being, of course, that any sustained recovery in labor should eventually lead to an uptick in consumer incomes and spending, helping to spur final demand. At the same time, Fed watchers will be monitoring developments in the labor market to help time the eventual onset of its inevitable tightening cycle. For these reasons, labor indicators, especially the monthly employment report, will take on even greater significance for investors in the coming months.

I've decided, therefore, to present a crude, but fairly accurate model that not only provides a better understanding of the dynamics of the monthly employment report, but underscores the relationship between the change in payrolls and several other employment indicators.

The model consists of three variables -- the ADP employment report, the non-manufacturing ISM employment index and the four-week moving average of jobless claims. Over the past week, I have been harping on about how large unexpected swings in any of these reports could impact Friday's consensus forecast. And in this model, I have put these variables together, so in future, you can instantly gauge the potential impact yourself and harp right back.

The ADP employment survey: Importantly, the ADP employment data are always released two days before the employment report, whereas other indicators are not be available until after the report. In such cases, I suggest you input the consensus forecast. Why are the ADP data significant? They measure non-farm private-sector employment. ADP processes payments for one out of every six private employees, giving the data a significant amount of meat.

Jobless claims: For this application, we input the most recent four-week moving average, as jobless claims tend to be a forward-looking indicator for payrolls, so past data should not be ignored. It should be noted, however, that some research indicates that claims data are better at forecasting the changes in payrolls during a recession and lose some of that ability during times of economic expansion.