Loading...
Symbols:
Authors:

Beware of the Data Underlying the Economy

by: Jeff Miller April 10, 2007    
Jeff Miller

Jeff Miller

About this author: Author's firm:
Become a Contributor Submit an Article
  • Font Size:
  • Print
In the standard stuff they write about the monthly employment report, if the number is bad, it stands as reported. If it seems to be OK, it is time to look at the "internals" to find something that seems out of line. Worst of all, since the Fed and their 350 economists are too obtuse to see the truth, they will look at the erroneous data and make the wrong decisions about interest rates.

A bad number is bad -- and so is a good number.

This morning's WSJ provided an analysis by David Malpass, who has had a stellar multi-year record both predicting and explaining the economy:

Friday's solid report of the March labor environment showed more Americans working more hours at higher wages than ever before. Unemployment fell to 4.4%, instead of rising as many expected. The average hourly wage rose to $17.22. The strong jobs report, along with the January-February surge in spending, personal income and core inflation points to sturdy growth and an inflation problem -- not the recession and interest-rate cuts many predicted in early March.

Despite his performance and a superior grasp of the data underlying the economy, Malpass is accused of looking only at "recent" data.

Investors (and sports fans) should look carefully to find the real experts when making their decisions.

Reader Challenge: Find and report the last time that Abelson or Ritholtz took an employment report that seemed "weak" on the surface, but explored the internals to show hidden strength.

Related Articles

This article has 8 comments:

  •  
    Apr 10 10:19 AM
    If you believe the economy went from 1.5mm new homes annually to 808,000 new homes annually and ADDED 65,000 construction jobs, then these predictions make perfect sense. If you believe that gasoline went up 100%, houses up 50%, strip stakes over $10 a pound and bread over $2 a loaf, gold went from $300 to $665 an ounce and inflation is 2% annually, I'll give you that one also.

    These government statistics are produced by thousands of economists who think math is a substitute for judgement. I'll put my bets on Jim Rogers. These stocks are rediculously priced. Government statistics are rubbish.
    Reply
  •  
    Apr 10 03:25 PM
    Paul, I bet you're the first guy to trumpet bad numbers when they're released, but any good numbers are "rubbish," eh?
    Reply
  •  
    Apr 10 03:36 PM
    Seeking Alpha editors do a great job, but the version here loses a lot of what I was trying to do with this topic -- including having a little fun. Interested readers might find the original version (on my site) to be more amusing.

    Thanks - Jeff
    Reply
  •  
    Apr 10 04:20 PM
    A commentator once said, "do you really think they count the number of VCR's that enter the country?

    Consider this, every quarter they get a copy of every employers 941. That's the employer return for payroll taxes. All they have to do is add them up. 99% of them are E-filed by companies like paychex. Yet if you look at the treasury website, they're 9 months behind.

    They can't count the date that's fed to them accurately but they CAN count data thats estimated and
    seasonally adjusted!? haha It's friggin comical.

    The average government statistician couldn't find his own ass with a bloodhound.
    Reply
  •  
    Apr 10 07:01 PM
    wow Paul, I like your version better.
    We use "can't find his arse with his own hands" or can't find his left hand with his right...but bloodhound is stronger. Nice!

    Good numbers need to be taken with a healthy dose of skepticism - nothing wrong with it.
    Reply
  •  
    Apr 10 10:21 PM
    Paul -

    You are correct in saying that the government eventually goes back to compare actual numbers with the monthly NFP methodology. That was how they created the birth/death model. The numbers showed that job creation was greater than their estimates. This was because of the massive changes in new small companies.

    Had they ignored this, the results would be even worse. Does the model reflect current reality? We'll know in a few months. Can you suggest a better method? We think they are doing the best they can -- pretty much what we would do if called in as consultants.

    As I have written on my site, we usually do not have the data we want at the time we think we need it, so everyone is free to speculate on trends not captured by the methods.

    Having said this, those bearish on the economy have been completely wrong on job growth for several years, and remain in a state of denial about the actual numbers when they come in.

    We have no fixation on a method -- we would just like to make good investments. I'm sure that you share this objective.

    Jeff
    Reply
  •  
    Apr 11 11:22 AM
    THe primary reason that our "Payroll" data are misleading, is that they come from about 8000 big companies, many of which are not growing and others are downsizing...we hear about those every day! However, we have about 29 MILLION incorporated U.S. businesses, up from 4.6 million in 1982. Over 95% of these have less than 100 employees. But since 1982, we have been generariong about 600,000 to 800,000 new small businesses every year (about 25 million since 1982). This is where most of the new (uncounted) jobs are being produced...and why we have our monthly "oops report" corrections. Small businesses now account for a rising 2/3rd of thge U.S. GDP....large companies account for a declining 1/3rd of the GDP. Our Government data are decades out of sync with the real economy...
    Bruce
    Reply
  •  
    Apr 11 11:07 PM
    Bruce -

    I think you are right on target. The much-maligned birth/death model tries to correct for this, but it has a lag. The official count from state agencies also has a lag when the benchmark revisions are done.

    Somehow those who are in denial about job growth use these final revisions -- those that reflect the most accurate data from actual counts and small businesses -- to suggest that the entire process is "fixed".

    Jeff
    Reply
More by Jeff Miller

Investment Resources
Money Management 

Research 

Seeking Alpha provides these ads as a service to its leading contributors
Articles on related themes
Economy 

Bonds 

Dollar/Currencies 

Earnings