"There are three kinds of lies: lies, damned lies, and statistics."~ Benjamin Disraeli, later popularized by Mark Twain.
As of October, the unemployment rate has risen to 6.5%, higher than the peak in the last recession and the highest rate since 1994. While the 6.5% rate has increased markedly from its low set two years ago, given the current dynamic in the economy, that rate still doesn’t seem to pass the smell test and “feels” much higher. For more than two decades the relationship between the unemployment rate and the duration of unemployment (median weeks unemployed) showed a very close fit, with a 95% correlation. However, starting in 1994 the two series diverged with the correlation falling to 65%.
Figure 1
Source: Bureau of Labor Statistics
Figure 2
Source: Bureau of Labor Statistics
What happened was that the Bureau of Labor Statistics (BLS) in 1994 redefined how they measure unemployment, excluding discouraged workers. As a result of that and other changes, historical unemployment measure relationships began to diverge markedly. For example, the figure below shows the U3 unemployment rate (most widely used and quoted measure), the U6 unemployment rate, as well as two measures of unemployment duration. Since the early 1990s, the two duration measures diverged significantly from the U3 unemployment rate and more closely tracked the broader U6 unemployment rate. The relationship between the median weeks unemployed and the U3 unemployment rate diverged significantly since 1994, and the spread between them tracks with the spread between the U6 and U3 unemployment rates as seen below.
Figure 3
Source: Bureau of Labor Statistics
Figure 4
Source: Bureau of Labor Statistics
Because the way the BLS calculates the unemployment rate has changed, the widely used U3 rate cannot be used for historical comparisons. Since the median weeks of unemployment rate displayed a close relationship with the old measure of unemployment it likely represents a truer picture of the real state of unemployment than revised U3. Currently, the median weeks of unemployment are 9.4 weeks and would correlate to a 9.4% unemployment rate for the old U3 measure, which is almost 3% higher than the reported revised U3 rate and a level of unemployment not seen since the 1981 recession! While a 9.4% unemployment rate is alarming, there’s something even more disturbing, and that is the rate is likely to head much higher heading into 2009. The Conference Board’s “job’s hard to get” index provides a useful leading indicator for the unemployment rate with a one-year lead time, and it is still rising.
Figure 5
Source: BLS/The Conference Board
Not only is the unemployment rate set to rise significantly heading into next year, we are either already at or close to setting records for the number of workers working part-time not by choice as employment conditions erode rapidly. Those who are working part-time for economic reasons jumped to 6.70 million last month, close to the 6.86 million record in 1982. Those who are working part-time due to slack business conditions reached a new record of 4.73 million. While these numbers are alarming, they are absolute numbers and don’t take into account a growing workforce. Taking the absolute numbers of workers working part-time as a percent of total employment shows an alarming trend, though still below record levels seen over the past half-century.
Figure 6
Source: Bureau of Labor Statistics
Figure 7
Source: Bureau of Labor Statistics
Summing up the three categories to measure the total number of part-time workers shows that we are nearing the all-time high of 13.3 million people working part-time because there are not full employment opportunities available. On a relative basis as a percentage of total employment, we are still significantly below the peak of 15.2% set in 1982.
Figure 8
Source: Bureau of Labor Statistics
Figure 9
Source: Bureau of Labor Statistics
Records are meant to be broken and it is likely that unemployment measures in this recession will break those of the last half-century and come second only to the Great Depression. The ISM’s Purchasing Managers Index plummeted in October to 38.9, the lowest reading since the early 1981 recession, and is likely to head much lower in the months ahead based on leading economic indicators. The Economic Cycle Research Institute’s (ECRI) Weekly Leading Indicator Index (WLI) was released today and showed a -29.2% smoothed annualized growth rate, the lowest in the history of the index, even worse than the 1970s and 1980s recessions. The WLI leads the year-over-year percent change in GDP by three quarters, so we can expect significant economic weakness in 2009, and likely further deterioration in unemployment as well.
Figure 10
Source: Economic Cycle Research Institute/BEA
Market Observations
There are a few more topics I’d like to cover but to keep this WrapUp “relatively” short, I’ll simply touch on them below. First, there is a price to pay for all of the Fed’s and the Treasury’s actions, and that appears to be the insurance cost of protecting against a default on U.S. debt. The price of Euro-denominated credit default swaps (CDS) of 5-Yr UST’s continues to climb to new records, a development noted recently in Barron’s.
Uncle Sam's Credit Line Running Out? (11/11/08)
Trillions are no hyperbole. The Treasury is set to borrow 0 billion in the current quarter alone and 8 billion in the first quarter of 2009. "Near-term pressures on Treasury finances are much more intense than we had thought," Goldman Sachs economists commented when the government announced its borrowing projections last week.
It may finally be catching up with Uncle Sam. That's what the yield curve may be whispering. But some economists are too deaf, or dumb, to get it.
The yield curve simply is the graph of Treasury yields of increasing maturities, starting from one-month bills to 30-year bonds. The slope of the line typically is ascending — positive in math terms — because investors would want more to tie up their money for longer periods, all else being equal. Which it never is.
If they expect yields to rise in the future, they'll want a bigger premium to commit to longer maturities. Otherwise, they'd rather stay short and wait for more generous yields later on. Conversely, if they think rates will fall, investors will want to lock in today's yields for a longer period.
The Treasury yield curve—from two to 10 years, which is how the bond market tracks it—has rarely been steeper. The spread is up to 250 basis points (2.5 percentage points, a level matched only in the past quarter century in 2002 and 1992, at the trough of economic cycles.
Based on a simplistic reading of that history and the Cliff Notes version of theory, one economist whose main area of expertise is to get quoted by reporters even less knowledgeable than he, asserts such a steep yield curve typically reflects investors' anticipation of economic recovery. Never mind that the yield curve has steepened as the economy has worsened and prospects for recovery have diminished. Like the Bourbons, the French royal family up to the Revolution, he learns nothing and forgets nothing.
As with so much other things, something else is happening this year.
The steepening of the Treasury yield curve has been accompanied by an increase in the cost of insuring against default by the U.S. Treasury. It may come as a shock, but there are credit-default swaps on the U.S. government and they have become more expensive — in tandem with an increase in the spread between two- and 10-year notes.
Figure 11
Source: Bloomberg
Figure 12. Euro-denominated 5-Yr CDS as of 11/26/08
Source: Bloomberg
To end on a lighter, more positive note, it looks like we are going to get the long and overdue rally. The VIX index broke its 50-day moving average and it looks like we may see a short to intermediate end to the unwinding of the Yen carry trade. The Euro/Yen exchange rate has gone through a 61.8% Fibonacci retracement, a frequent stopping point for a decline. Both a decline in the VIX and further strengthening in world currencies relative to the Yen will point towards further market advances ahead. Potential targets for the S&P 500 would be the 50-day moving average (~ 970), November high’s (1007.51), or the 50% Fibonacci retracement level (1029.03) from the August highs.
Figure 13
Source: Stockcharts.com
Figure 14
Source: Stockcharts.com
Figure 15
Source: Stockcharts.com
While a rally in the markets is something to give thanks for, I still feel rallies should be used to increase defensive positioning as the credit crisis and economic outlook are still very powerful forces to be reckoned with. Last week’s Observation (11.19.2008) focused on the important concept of risk management and highlighted the Bloomberg Financial Conditions Index (FCI) as a barometer for risk. While a current reading of 6.74 is a welcome event from the 10 standard deviation event seen in October, the current reading is more than twice the worst level of the 2000-2002 bear market and investors should continue to remain highly defensive and give thanks for any market strength to sell into.
Figure 16
Source: Bloomberg
HAPPY THANKSGIVING!
