Thursday, January 31, 2008

Tomorrow's job data

Payroll and unemployment numbers come out tomorrow. I've been watching the Birth/Death model for months now. This model is meant to adjust employment numbers to take into account the Birth and Death of businesses. The problem is that statistical forecasting is much better at predicting the trend going forward than it is the turning point. If the economy turns south, the BD numbers will make the situation appear rosier than it would be otherwise. Mish, Barry Ritholtz, and John Crudele have all expressed concerns about the numbers. Briefing.com shows the consensus around 70,000-80,000 compared to a prior number of about 18,000. I'm agreeing with Barry and John that the number will probably be significantly lower. Maybe as much as 25-50% lower. The reason, explained by each, is that every year there is an annual rebalancing of the Birth/Death Model to recalibrate their estimations. If there is a significant deviation of the model with reality, the model will be adjusted to take that into account. In a time when advance GDP in Q4 came in at .6% and Q3 was only up due to better than expected inventory numbers, it's likely that there will be a significant adjustment.

From the short-term perspective it will be curious how the market reacts. If they are not aware of this information (traders read the Post for sports not finance), they will think the economy is worse than expected and play up hopes for a rate cut. However, they will have to weight rate cuts with a shitty economy. The person who follows the Birth/Death numbers already knows that the intermediate outlook for the economy is absolutely awful and the numbers have made it appear much better and should already be short (or in bonds) in that time frame. In the shorter-time frame, it will be easier to just take advantage of those who don't read the Birth/Death numbers. For the short-term, Gold will be a buy no matter what the market thinks will happen.