An Urban's Rural View

Jerked Around by the Numbers

Urban C Lehner
By  Urban C Lehner , Editor Emeritus
Connect with Urban:

The crusty Marine veteran who ran the Washington bureau copy desk of the Wall Street Journal in the 1970s sometimes lost patience with his twentyish economics reporter -- me.

My stories on the government's economic indicators -- unemployment, factory orders, I covered them all -- would occasionally quote economists saying, "Don't read too much into a single period's data. It takes several reports to establish a trend. And remember the numbers are subject to revision."

The copy-desk chief considered that waffling. "There's always somebody telling you the numbers don't mean what they say," he grumbled. Readers, he argued, make decisions using the information in front of them today. They don't wait for economists to decide months or years later what had really been going on.

He had a point, of course. Many readers do indeed make decisions relying on statistical indicators, even those who realize the next government report might contradict this one, forcing them to reverse course. Some traders buy or sell the instant a headline crosses the wires, then turn around and make the opposite trade later in the day when the full story is out.

In essence, the copy-desk chief was telling me, "You may be right, but you have to accept that this is the way the world works." To which, in essence, I was responding, "If by including these caveats I can convince someone who doesn't have to make a decision using these numbers not to -- to wait instead until he understands what's really going on -- I will have performed a service."

It took another 25 years for me to meet Darin Newsom and realize that a similar service could and should be performed with regard to the government's agricultural statistics. Yes, many commodity traders will continue to trade -- or, these days, program their computers to trade -- based on USDA reports. That's the way the world works.

But in the view of the DTN analyst team led by Darin, hedgers are better served basing their positions on an understanding of what's really going on. Much of that kind of understanding comes from a sophisticated reading of the market's own signals. That's what DTN's analysts try to provide.

For a case study in what can happen when decisions are based on the latest economic indicators, consider the Federal Reserve's monetary policy. For some time now the Fed has wanted to gradually move interest rates from near-zero back to something approaching normal. It has told the markets it will be "data-dependent" in deciding whether to raise rates at a particular moment.

But the economic data has been jumping around, and as a result the Fed has lurched from preparing the market for another increase to reassuring the market that the time is not yet ripe. It has done this not once but twice since late last year. Most recently, it had to back off from a threat to raise rates in June when the Labor Department reported that the economy created only 38,000 jobs in May.

It's easy to blame the government's statistical mills for the choppiness of the data, but in my experience the bureaucrats are competent and honest. They do the best they can with the tools they've got, trying to make sense of an enormous and enormously complicated economy. Over time they eventually get things right. In a single period's report they sometimes tell you how confident they are in the accuracy of their work.

The Labor Department did that with the May jobs report. As the Wall Street Journal reported, "The Bureau of Labor Statistics is 90% sure that its conclusion of 38,000 jobs created is right to within 115,000 jobs, and so lies somewhere in the range of a loss of 77,000 jobs to the creation of a 153,000 jobs (…)."

Not only is this, to use the Journal's technical terminology, "a mind-blowing range," combined with the 90% confidence level, it means that one month out of every 10, the real number could be even more than 115,000 away from the reported number.

The Journal's reporter has obviously had the same kind of conversations with editors that I had back in the 1970s. He says the most popular answer to the question of what an investor should do about the shakiness of the data is "treat the figures as though they were gospel truth, on the basis that everyone else does. Pedants don't make money, at least in the short term."

The Federal Reserve has the markets to worry about, and markets are not pedants. So arguably the Fed has no choice but to pay attention to the data. But by depending on it as slavishly as this Fed has, the central bank risks being battered senseless by the fusillade of fluctuating statistics. For as the headline on the Journal story aptly put it, "You Know Almost Nothing About the Economy: Get Used to It."



To comment, please Log In or Join our Community .