Is anybody listening out there? Those ‘in the know’ expected 180,000 job losses. Some thought 150,000. Optimists thought job gains. In the end, we bled 283,000 jobs.

It was no surprise to me. The message has been clear in the past two Duke-CFO quarterly surveys. Companies are still in cutting mode. Employment is not going to significantly improve when we have 551,000 new claims for unemployment insurance. So what if the four-week moving average of claims has decreased by 6,250 to 548,000. We need some three handles to stabilize employment and we are way far from that territory.

What to Watch For

The following are very important for those of us that carefully track employment:

• Forget continuing claims. They are misleading. Any headline like “Good News, Continuing Claims for Unemployment are Down” is likely written by someone who doesn’t really understand the data and the Federal programs. Continuing claims might be lower because people roll off the standard unemployment program and hit extended benefits and the emergency program (EUC). You need to count correctly. Notice that the extended benefits and EUC are reported with a two week lag. The most recent data showed that the number of people on emergency programs increased by 99,832 to 3,275,213.
• Extended benefits and EUC are important to watch because we have a problem with the duration of unemployment. It is taking longer to find jobs and this is another sign that the economy is not coming around.
• Watch the average number of hours worked in a week. The number of hours worked has been decreasing. People are sometimes voluntarily working fewer hours to save their jobs (or fellow workers’ jobs). Working fewer hours is a “shadow” unemployment that is not counted in the official numbers.
• I like the Bureau of Labor Statistics measure “U6”. This measure looks at total unemployment by considering unemployed, “marginally attached” workers as well as people working part-time that really want to work full time. This rate is a staggering 17%. It has never been this high.
• Initial claims as mentioned above need to be in the 300,000 range to stabilize employment. We are nowhere near that.
• The scale of the unemployment problem is important because we are approaching no-man’s land. What I mean is all the data to analyze the impact of unemployment comes after 1947. We are already in a recession situation worse than any other since WWII. Yes, this is not the Great Depression – but we don’t really have good data on the Great Depression. It is a challenge to quantitatively analyze some problems (I have an example below). To be clear here, we have lost 7.2 million jobs during this recession. We will lose more. How about this comparison? If we lose an additional 1.7 million jobs, i.e. go to 8.9 million losses from 7.2 million, the loss of jobs on a population adjusted basis will equal the losses in the past four recessions — and that includes the nasty double dip of the early 1980s.

What Worries Me the Most

Suppose you statistically examine the relationship between unemployment and mortgage defaults. There is a moderate positive association as you might expect (higher unemployment leads to higher defaults). Indeed, this moderate positive association forms the basis for the stress tests that banks have (finally) done.

What if this model is flawed? What if the financial institutions have underestimated the number of defaults? Here are the reasons why the models could be wrong:

• The estimates are largely based on mild recessions. This is not a mild recession. We have not seen anything like it in all of the available data. Essentially, we are extrapolating outside of historical experience. Usually when this occurs, we explicitly account for the extra uncertainty by being more conservative. I have my doubts that the financial institutions have factored in this extra uncertainty.
• While first point is important, the second one is more important. There is a critical omitted variable in the analysis. People statistically analyze default versus unemployment — but you must take housing values into account. If your house is worth less than what is owed on the mortgage, you are far more likely to default than if your house is worth more than the mortgage. In most of the data used in the standard analysis, houses are worth more than the loans. In other words, the fact that potentially 50% of houses with mortgages are under water, greatly increases the sensitivity of default to unemployment. This suggests that there is much more downside than financial institutions have prepared for. It suggests a second wave of credit crisis.
• Even the financial institutions know that defaults have not peaked yet. People are still holding on. Expect a surge in defaults, foreclosures and bankruptcies to happen next year.
• The social stigma attached to defaulting on your mortgage is gone. Keeping up the Jones’ means you get to default too. Anyways, those banks took all that government money making it easier for them to be the bad guys.

Other Tidbits

• As predicted, the Cash for Clunkers simply shifted car purchases. GM sales down 42%. In the end, the problems in the auto industry remain. Any bets on Chapter 22?
• There is a lot of push back on basic financial regulatory initiatives – like the requirement to offer vanilla mortgages. A lot of the push back is from small financial institutions who cannot compete in such a market against the likes of BAC and Citi. The real issue here is why do we need 7,000 banks in the U.S.? It would be far better for both customers and the economy to have a massive shakeout. In addition, it is completely dysfunctional to have different banking regulations depending upon the state that the bank resides in.
• The FDIC is bankrupt. For the last year, I have said that there are 1,000 banks that need to go down. Now we are in the odd situation where the FDIC cannot close down so many banks because they simply don’t have the funds.

Follow me on Twitter.

See below my monthly employment graph that standardizes the job losses (based on the size of the labor force) across different recessions.