John Hussman has been predicting precisely this kind of data divergence in anticipation of the next stock market correction.
Morgan Stanley's David Greenlaw addresses this in a recent note to clients:
There has been an unusual divergence in the data flow in the US. Last week, our tracking estimate of Q1 GDP slipped all the way from +2.2% to +1.0% in response to disappointing reports on durable goods, construction spending and personal consumption. At the same time, jobless claims continued to improve, consumer sentiment showed a decent pickup, chain store reports were positive, and motor vehicle sales surprised to the upside. The sharp divergence between the GDP arithmetic and better-than-expected performance from a range of indicators is unusual, but not unheard of. Most importantly, the economy's underlying growth trend still seems to be in the neighborhood +2.0% – consistent with both the +3.0% outcome seen in Q4 and a +1.0% tracking estimate for Q1.
Other economists who have had to make some big revisions to their GDP estimates include Goldman's Jan Hatzius who last week revised his Q1 GDP estimate from 2.4 percent to 2.3 percent, then again to 2.0 percent. Bank of America also recently cut its estimate from 2.2 percent to 1.8 percent.
