Photo: AP Images
This post is excerpted from the permanent Employment Charts page, which includes numerous additional charts and analysis on key employment metrics.
The labour Department reported that the seasonally adjusted (SA) representation of first time claims for unemployment rose by 38,000 to 368,000 from a revised 330,000 (was 335,000) in the advance report for the week ended January 26, 2013.
The number was far worse than the consensus median estimate of 345,000 reported by major business media outlets. The business of seasonal adjustment remains a haphazard and arbitrary process that always yields a fictitious number. For a third consecutive week the problem was made worse by turn of the year calendar factors further exacerbated by leap year last year. Professional pundits reported last week that they thought the numbers would return to “normal” at the end of January. They moved in that direction, although full “normalization” would happen in March when the leap year, 366 day problem, goes away. Even when the comparisons “normalize,” however, the seasonal adjustment problem will not go away. That the SA number is fictitious, arbitrary, and haphazard is a permanent feature of the data.
The headline seasonally adjusted data is the only data the media reports but the Department of labour (DOL) also reports the not seasonally adjusted data (NSA). The DOL said in today’s press release, “The advance number of actual initial claims under state programs, unadjusted, totaled 366,596 in the week ending January 26, a decrease of 70,429 from the previous week. There were 422,287 initial claims in the comparable week in 2012.” [Added emphasis mine]
This year’s filings represented an decrease of 13% over the referenced week last year. Due to the calendar factors, the current week would more correctly be compared with the week ended January 21, 2012. Using that date the decline was 49,000 or -11.8%, not a material difference. The year to year comparisons are beginning to normalize in the actual, unadjusted data, but not in the seasonally adjusted data, which is giving an unusually misleading picture this week, particularly in terms of the analysts’ expectations game. The problem this week is that the economists were actually closer to being correct this week. The SA data was just way off the mark in correctly representing the trend.
Claims always peak in the first full week ended in January. This year, that was the week which ended on January 12. Last year it was the week ended January 7. Because the government and mainstream analysts make the “year-to-year” comparison versus 52 weeks ago, not exactly a year ago, that’s a problem this year, partly due to the fact that last year was a leap year with 366 days or 52.29 weeks . Therefore, comparing the January 12 week this year with the January 14, 2012 week was not an “apples to apples” comparison, nor are the following weeks. In this case, the “like” week for the January 12 week this year was the one ended January 7, 2012.
Whereas they always rise sharply the first week of the year, claims always fall sharply in the second through fourth weeks of January. The current week is 3 weeks after the first full week this year. Last year the like week comparison would be the one ended January 21, 2012. Using that date, the drop of 11.8% is consistent with the trend this series has been on for the past couple of years.
So is the drop of 13% in the government reported NSA data cited above, so that using the January 28, 2012 week as the benchmark week would not result in a misleading comparison in the actual, NSA data, whereas the SA data is misleading. Regardless of which week the current week is compared with in the actual NSA data, this week’s number is another good one in a long trend of steady improvement.
Note: The DOL specifically warns that this is an advance number and states that not seasonally adjusted numbers are the actual number of claimants from summed state claims data. The advance number is virtually always adjusted upward the following week because interstate claims from many states are not included in the advance number. The final number is usually 2,000 to 4,000 higher than the advance estimate. I adjust for this in analysing the data.
I adjusted this week’s reported number up by 1,000, which is consistent with last week’s revision. The adjusted number that I used in the data calculations is 368,000, rounded. Don’t try to read anything into the fact that this is the same as the SA number. It is just a meaningless, random coincidence. It simply means that the adjustment factor, which can vary by as much as 10% or more for the same week from year to year (see below), was 1.ooo. That rarely happens, but it did this week.
Note: To avoid the confusion inherent in the fictitious SA data, I analyse the actual numbers of claims (NSA). It is a simple matter to extract the trend from the actual data and compare the latest week’s actual performance to the trend, to last year, and to the average performance for the week over the prior 10 years. It’s easy to see graphically whether the trend is accelerating, decelerating, or about the same.
The week to week change of a decrease of 69,000 in the NSA number was not as good as usual for the third full week in January. Over the previous 10 years the comparable week has always had big declines. The average change for the 10 years from 2003 to 2012 was a decrease of approximately 96,000. In 2012 it was 108,000 and in 2011 it was 64,000. This difference is within the usual variance, and probably not significant.
From mid 2010 through mid October 2012 the annual rate of change in initial claims had ranged from -3% to -20% every week, with a couple of temporary minor exceptions, including the Superstorm Sandy surge. Since mid 2011 the annual rate of change was within a couple of per cent of -10% in most weeks. The trend was remarkably consistent.
A second trend has become visible on the 52 week rate of change graph (bottom of chart below). It shows a channel of slightly higher lows and higher highs indicating a slowing rate of improvement as the trend moved toward zero year to year change. The “like-to-like” week, annual rate of change of -11.8% is near the middle of these trends.
Normally, I would expect some moderation in the rate of improvement in the weeks and months ahead. Further reductions in the number of new claims should be much more difficult to achieve going forward as the year to year comparisons become tougher. However, the Fed’s money printing could be stimulating bubble dynamics, which could cause employment growth rates to increase. The markets may cheer this, but it will be artificial, sustainable only as long as the Fed can sustain its money printing operations.
Photo: Wall Street Examiner
Plotted on an inverse scale, the correlation of the trend of claims with the trend of stock prices over the longer term is strong, while allowing for wide intermediate term swings in stock prices. Both trends are largely driven by the Fed’s operations with Primary Dealers (covered weekly in the Professional Edition Fed Report; See also The Conomy Game, a free report).
The chart below suggests that the market trend is overbought at approximately 1500, assuming that the trend in claims remains relatively constant. This has been a long standing projection. Now that the market has reached this parameter, the question is whether additional Fed money printing will cause those trends to tilt more steeply upward. The year to year trendline in claims has been remarkably consistent. If stocks break out, unless the claims trend begins to show consistent evidence of acceleration, stocks would become more overbought versus that trend, and eventually vulnerable to a big decline.
Photo: Wall Street Examiner
Under the current QE regime, the Fed’s balance sheet will grow by a 38% annual rate, sending lots of cash through the market, with some of it trickling into the economy for the duration of the program. This is unlike 2011 when the market became extended relative to the unemployment claims trend. Then, the Fed was simultaneously ending QE2, thus starving the monster of its lifeblood. As a result, the market pulled back sharply after reaching the top of the channel This year, the Fed is intent on fattening the bull. That would allow the S&P to bump along the top of the channel as long as the jobs trend stays intact. However, any breakout in stock prices without a complimentary acceleration in the the improving trend of claims would lead to a dangerous over-extension in stock prices.
Some bubble jobs will likely be created as the Fed pumps $85 billion per month net into the financial markets. However, the inflation that should accompany the money printing, whether in asset prices, commodities, or in consumer prices, should eventually force the Fed to stop QE. At that point the markets and economy will deal with the hangover from the program. The greater the extension of stock prices above the trend of claims on this chart, the greater the risk of a major correction, crash, or bear market.
[I cover the technical side of the market in the Professional Edition Daily Market Updates.]
Permanent Employment Charts page