Calculated Risk has the following revealing graph:
Click on graph for larger image.
The ISM Manufacturing Employment Index for February was at a level of 56.1. Using the linear relationship shown in the graph that would correlate with an increase in employment in the manufacturing sector in the BLS (Bureau of Labor Statistics) data. But note that the correlation coefficient (R-squared) is less than 0.5. A value of 1.0 would indicate perfect correlation with all data points on the line. A value of 0 indicates absolutely no correlation (random scatter). The value shown indicates that there is a correlation, but it is poor. The data is a little closer to pure random scatter than it is to a perfect straight line.
The other factor that dampens my enthusiasm is the fact that before 1988 (red points), the BLS employment tended to be on the high side of the line when the ISM Index was in the mid 50s. In the last 20 years the data points (blue) are almost all on the low side of the line. In fact almost half appear to be negative for change in the BLS data.
So we have some good news from ISM, but statistically it may not translate into good news for the BLS numbers out later this week. Of course, one data point is never a statistically significant data set, so the Friday data from the BLS can be anywhere.
Disclosure: No stocks mentioned.