It would be good to go back and review Predicting GDP With Two Variables. Keep in mind that I generally don’t see anything actionable in modeling GDP, jobless claims, or other such stuff, and don’t use that type of modeling in my trading. My point in showing these models is to demonstrate that the vast majority of econopundits are outsmarting themselves – not that that’s a difficult proposition – and that a bug with access to data and some skilz can do a better job in 30 minutes or less.
The initial take was a prediction of next Calendar Year’s GDP change using the Consumer Price Index: All Urban Consumers, All Items and Civilian Unemployment Rate from the end of the second quarter of the previous year, that is, using final June xx data to predict final change over calendar year xx+1. The model looked promising, and showed CPI increase as a drag on future GDP and high unemployment as an engine for higher future GDP, both of which make sense from an Austrian Business Cycle perspective. Based on that data, 2008 was projected to be at trend (+3.3%) and above the projection that (would have been) made for 2007 (based on 2Q ‘06 data) of +2.7%.