Lies, Damned Lies, & Statistics

The incumbent president's political party has a vested interest in presenting a happy face to the voters, especially during an election year. It has several tools it can use to enhance the appearance it presents. One, of course, is the Federal Reserve. When the central bank buys Treasury bonds, it lowers interest rates and floods the economy with dollars to stimulate the economy.

As investors, we have a duty to ourselves to form a valid opinion about the economy -- one based on reality, rather than the subtle things that politicians and trade groups do to manage the news.

What do I mean by news management? Consider this example from the National Association of Realtors. Granted, the NAR is not a government agency. But this example of data manipulation will serve as a model before we address direct government manipulation later in the post.

The NAR admitted that it had erred in reporting sales of existing homes, and it revised its figures all the way back to January 2007. This chart plots the year-over-year growth in sales, along with the change in the average price of a house.

The change in price was reported correctly, but the number of houses changing hands was dramatically higher in the initial reports. This was more than a casual error. On average, sales were reported 13% higher than they actually were. At one point in 2010, sales were reported to be down 35% from their high, when in fact they had fallen by half.

Clearly, it is in the best interest of the NAR to conceal the depth of the slowdown in sales from prospective buyers, who might delay a purchase because they see even lower prices ahead. It also scares off sellers, who realize they are likely to have difficulty finding a buyer.

But the practice of revising data is not limited to private organizations. We have been told that the national unemployment rate has been understated at 8.6%, because several hundred thousand unemployed people were removed from the labor pool. This is difficult to check, but here's an example that can be easily documented.

On December 8, the Department of Labor released the following note with its weekly employment data report (with emphasis added): "In the week ending December 3, the advance figure for seasonally adjusted initial claims was 381,000, a decrease of 23,000 from the previous week's revised figure of 404,000."

In the report from the previous week, the department had reported a seasonally adjusted total (a great source of data manipulation in itself) of 402,000 initial claims. That figure was revised higher to 404,000 -- a change of merely 0.5%. However, the report focuses the reader's attention on the 23,000-claim decrease, so there is an improvement of 8.7% in the number on which the public is asked to focus.

This would still not be a big deal if it were an occasional error made randomly, but it is not. It is a systematic error that has occurred in the same direction in 56 of the past 58 reports. The revision increased the number of newly unemployed by an average of 3,286 each week. The weekly figure was off on average by an amazing 57%.

The peaks in reported average change in initial jobless claims (solid blue line) represent a decrease in claims. Notice how the reports exaggerate the reduction in unemployed compared to the revised figures released each week.

Notice in particular how the data in May and June show unemployment to be decreasing while in fact the rate was increasing. This is not a unique government error. I have recorded similar errors in many economic indicators, and virtually all of them paint a rosier picture than reality can support.

Here's another example: the quarterly corporate profits reported by the Bureau of Economic Analysis. The blue dots represent the initial data, and the blue line represents the data after revision. Only one revision was higher in the last four years.

Here's one that is no surprise. Sixteen of the past 20 monthly reports of year-over-year changes in the Federal Housing Finance Agency's Home Price Index were revised lower. The index represents the activities of Fannie Mae and Freddie Mac.

Stock traders can profit from the rosy picture painted in an election year. But they must be prepared for the truth when it emerges, and they must resist the urge to be blinded by the promise of hope and change.

— Fred Goodman, a registered investment adviser and Certified Financial Planner, publishes MarketMonograph, a daily, Web-based subscription service specializing in technical stock market analysis and the application of economic indicators to market timing. You can reach him at fred@marketmonograph.com.