http://www.zerohedge.com/news/2013-01-10/revision-wipes-outs-surprising-december-philly-fed-surge
Revision Wipes Outs "Surprising" December Philly Fed Surge
On December 20, when we posted on the miraculous surge in the Philly Fed, offsetting the far weaker NY Fed data, we were left scratching our heads as the upwardly inflecting data made little sense in the context of broader data. To wit: "Three days ago the New York Fed released the December print Empire State index which showed a broad contraction across all key verticals. Today, in fine "keeping them baffled with bullshit" form, the Philly Fed swing precisely the other way, and despite expectations for a second consecutive negative print of -3 to be precise, up from -10.7 last month, the General Business Activity indicator printed at 8.1, the highest print since April, with New Orders at 10.7, the highest since February, and Employment at 3.6, the highest since April. Naturally, the algos pretending to trade on news, took this news and ran futures higher..."
We also added, rather providently, "Needless to say, all economic data in the US at this point is completely meaningless, with regional distortions, seasonal adjustments, political pressures and overall central planning making a mockery of the US economic data apparatus."
Today, 20 days after the data release, we get the explanation for this very surprising jump, which naturally put the algos in a buying tizzy and sent the market higher by 1% (before it flash crashed late in the night) on the date of its release, as well as the latest validation of our skepticism, courtesy of the Philly Fed annual data revision. To wit: "Overall, the data revisions adjust the current year-end indexes moderately downward. There were more pronounced downward adjustments of the six-month expectations indexes over the last two months of 2012." And of course, it is the last data point - December - that matters most. So how does December looks like pre and post-revision? Well, it is self-explanatory: look at the chart and decide for yourselves - blue is original, red is revised.
This is the kind of "credible" data that drives headline scanning algos.
