The Case for Depression, Part 1: Unemployment

Includes: DIA, SPY
by: Moses Kim

It's pretty intuitive that recovery isn't possible without sustained improvements in the rate of unemployment. The most recent unemployment reading for May was 9.4%--bad yes, but a far cry from the 25% levels and soup lines we saw during the Great Depression.

Now let's dig a bit deeper into the data. The "official" unemployment rate that gets all the headlines is the U-3 rate. The extent that the U-3 captures what the average person would define as unemployment is dubious. Three important categories are left out: (1) people not actively seeking work, or marginally attached workers, (2) discouraged workers, or people who have given up looking for a job because of the perceived weakness of the job market, and (3) part-time workers who want full-time work but can't find it. These 3 additional categories make up the broader calculation of unemployment, or U-6. U-6 is currently at 16.4%.

Click to enlarge

Further, improvements in the 2nd derivative decline in unemployment are overstated. In any normally distributed curve, the highest rate of change will naturally occur at the inflection point. Even if we have reached the inflection point in unemployment, which I doubt, there are still many more months of job losses ahead.

Now lets take a look at the BLS' "Birth/Death model". The Birth/Death model adds or subtracts to the headline unemployment number every month based on the rationale that a certain number of businesses are created and destroyed over a given period of time. However, since the created and destroyed jobs that result can not be adequately captured in current payroll data, estimates on net job gains/losses are made using econometric models based on historical benchmarks. Since our economy has generally been quite robust, estimates are on the optimistic side and jobs are regularly added. However, in economic downturns--such as the current recession -- the models will severely understate unemployment. For example, 112,000 jobs were magically added the past 4 months in construction amidst the biggest real estate downturn in history. A total of 694,000 jobs were conjured up the past 4 months.

Click to enlarge

To add, April, May, and June are generally strong seasonal months for the birth/death model- which explains in part the recent "improvement" in unemployment. As you can see, the unemployment picture is a lot worse than advertised.

For a number of reasons, the present unemployment situation is not merely a temporary aberration. My reasoning lies in the way our economy is structured, with an overreliance on credit, which I will cover in Part 2 of "The Case for Depression".