Just two days ago we pointed out [7] that Janet Yellen's San Francisco Fed has decided it's time to take back the media spotlight from its Atlanta counterpart which, thanks to an incredibly prescient, real-time GDP forecasting machine, has managed to make the financial news quite often after we first highlighted [8] the "nowcast" back in March.
In sum, the San Fran Fed decided that the weak Q1 GDP data simply needed to be seasonally adjusted (again) in order to make it ...well, less weak.
The official estimate of real GDP growth for the first three months of 2015 was shockingly weak. However, such estimates in the past appear to have understated first-quarter growth fairly consistently, even though they are adjusted to try to account for seasonal patterns. Applying a second round of seasonal adjustment corrects this residual seasonality. After this correction, aggregate output grew much faster in the first quarter than reported.
Got that? There was still some "residual seasonality" (i.e. the data still looked weak) in the Q1 print after the first round of seasonal adjustments, so in order to "correct" things, a second round of seasonal adjustments needs to be applied, after which the new figures should show that the economy did not in fact flatline in the first three months of the year. Of course if the numbers still don't come out looking the way you want them, you can always rinse and repeat. As we put it two days ago:
And if the double seasonally adjusted data doesn't work? Why triple adjust it, then quadruple adjust it, until you get precisely the goalseeked number you want, as US economic "data" promptly devolve to a level of ridiculousness that will make even the Chinese Department of Truth turn green with envy.
Sure enough, just moments ago, CNBC's Steve Liesman (who else) reports that the double seasonal adjustments are indeed in the works.
Here's Liesman himself [9]:
The government agency charged with calculating the nation's growth rate is acknowledging problems with its numbers and pledging a series of fixes over the next several months.
In a statement to CNBC, the Bureau of Economic Analysis said it's "aware of issues" in it gross domestic product data and "is developing methods to address what it has found."
The BEA statement comes after CNBC, in a detailed report in April, showed that first quarter GDP data has been weaker than the other three-quarters for the past 30 years and substantially weaker in the past five.
Several economists, including researchers at the San Francisco and Philadelphia Federal Reserve banks and many Wall Street economists, have since confirmed CNBC's findings. Many attribute the problem to what is known as "residual seasonality" which are seasonal patterns that remain in the data even though it is already adjusted for seasonal variations.
Nicole Mayerhauser, chief of BEA's national income and wealth division, which oversees the GDP report, said in the statement that the agency has identified several sources of trouble in the data, including federal defense service spending. Mayerhauser said initial research has shown this category of spending to be generally lower in the first and the fourth quarters. The BEA will also be adjusting "certain inventory investment series" that have not previously been seasonally adjusted. In addition, the BEA will provide a more intensive seasonal adjustment quarterly service spending data.
There you have it. Full-statistics-massaging-goalseeked-retard, and now Janet Yellen can completely ignore the miserable GDP print on her way to hiking rates sooner rather then later, plunging the economy into a recessionary tailspin which can then promptly be used to justify QE4.
In the end, it appears as though the BEA is about to replace all of its goalseeking excel models with this:
[10]
