PPI Follow Up

This may be a bit wonky, but follow it thorough to the end.

Yesterday, we discussed how you can avoid the usual seasonal adjustments and other shenanigans by using year-over-year prices. Those numbers reveal that inflation is far from benign.

Rereading the BLS release on the train on the way home last night, I noticed another aberration: In October 2007, BLS reports that energy prices were down 0.8% in October.

As absurd as that appears on its face, there is actually a simple explanation for this. The methodology for measuring PPI contains a simple and dramatic flaw: It measures prices on a single day of the month. BLS samples for energy prices on the Tuesday of the week that contains the 13th of the month. In other words, BLS’s methodology essentially ignores all energy prices paid EXCEPT FOR ONE DAY OF THE MONTH.

Most of the energy price increases occurred later in In October. The so-called monthly measure of Producer prices failed to pick up any changes in prices for the 2nd half of the month.

This methodological failure results in some spectacularly absurd results. To wit: BLS has gasoline prices down 3.1%, and energy prices down 3.64% since July. Even as Oil rallied from ~$75 to the mid-90s since July 9.  Over the
same period, gasoline has rallied from $1.95 to $2.35; heating oil has
rallied from $2.15 to $2.55; natural gas has fallen from $8.50 to


As we have so laboriously explained, there is no grand conspiracy. Rather, the BLS data outputs are merely the result of lousy statistical methods. This helps to explain why the official data manages to not see the inflation that any sentient beings with functioning visual orbs can’t miss.

CPI data is out at 8:30 am. I would tell you what the consensus expectations are, but by now you should you have already figured out that they are totally irrelevant to any reasonable approximation of reality . . .

Print Friendly, PDF & Email

What's been said:

Discussions found on the web:
  1. Hal commented on Nov 15

    Uh–what happened to oil prices on November 13th?

  2. Jay Weinstein commented on Nov 15

    BR –your excellent work on misleading government data is what first got me hooked on the site, and continues to be the most valuable analysis on here IMO.

    thanks as always!

  3. Jay Weinstein commented on Nov 15

    But don’t ever semi-apologize for being “a bit wonky”! It is such wonkiness that is critical to being a good investor.

    Look at Calculated Risk–some of the wonkiest stuff ever, and also probably the best stuff on the internet.

  4. mhm commented on Nov 15

    Off topic, but NYSE is down to just five specialists firms now…

    On Thursday, Van der Moolen Holding NV said it will close its NYSE market-making unit VDM Specialists USA after losing 11.3 million euros in the first nine months of the year. The announcement came just a day after SIG Specialists Inc., a unit of Susquehanna International Group LLP, said it too would stop acting as a stock specialist.

    The departures follow multiple rounds of layoffs that have cost hundreds of jobs, and will leave the NYSE with just five specialists. They include units of Bank of America Corp. (BAC), Goldman Sachs Group Inc. (GS) and Bear Stearns Cos. (BSC), as well as LaBranche & Co. (LAB) and Kellogg Specialist Group.

  5. ferd mertz commented on Nov 15

    no grand conspiracy , but way back in the clinton years, greenspan testified to congress that inflation measures overstated inflation. as john williams points out, BLS subsequently changed their methodology to skew their reporting downwards. works great for minimizing entitlement spending with cost of living adjustments.

  6. Stuart commented on Nov 15

    Yes, perhaps no grandiose conspiracy, but policy makers are certainly cognizant of the impact of understating CPI on COLA payments to pensioners. No doubt policy makers openly discuss the impact of their released CPI figures on what that means for COLA to pensioned income. This is a large reason for the consistent bias towards understating inflationary expectations.

  7. Old Ari commented on Nov 15

    Perhaps they should release two figures, one inflation, two cost of living, they seem to be different,one concerns increase in money, while the other, retail items.

  8. mack macdaniel commented on Nov 15

    “…any sentient beings with functioning visual orbs can’t miss.”


    thanks Barry. that was funny.

  9. OldVet commented on Nov 15

    Wacky market moves this a.m. attributed to tame CPI reading, in which higher energy/food prices were offset by lower Owner Equivalent Rent prices.

    Rents were falling as demand for rentals increased due to mortgage foreclosures? Really? And while mortgage payments are resetting upwards for many Americans with ARM’s? WTF.

  10. SINGER commented on Nov 15

    Like you always say…The information is out there…it just takes a bit of digging into and thinking about to uncover the fallacy…

  11. nemo commented on Nov 15

    Why don’t they average over the month? Any reasonably bright 12-year-old could have figured that out.

  12. GRG commented on Nov 15

    Also, why on earth do news organizations not do any research of their own? Looking at all the NYC papers, they talk about inflation being tame… basically repeating word for word the data they get from the government.

    How is it that blogs have become more reliable with information than the old, venerable news organizations?

  13. deltaverde commented on Nov 15

    The federal government purposely using bad methodology to mislead the populace? In what way is that not a grand conspiracy?

  14. muckdog commented on Nov 15

    BR, insulting the bond market?

  15. David Sternfeld commented on Nov 15

    slightly OT: How can anyone be surprised with ANY economic info published by these political hacks? They are directed by the same incompetents who’ve for five years have been telling us the great progress we’ve been experiencing in Iraq. (snark)

  16. TimW commented on Nov 15

    DeltaVerde and others.. the story of a grand political conspiracy is unlikely, in my opinion. I work for an economic indicator (not BLS, so I can’t speak for them). It’s unlikely that your typical agency working on the methodology for an economic survey would just willingly change methodology for some political appointee. There would have to be a good, statistical reason for the change… or I grant you some major incompetence or bamboozling going on.

    Where I am there would be institutional resistance to changes in methodology that was perceived to be politically motivated. First of all, the politicos that I deal with have very little knowledge of the technical details of the methodology that we use. Their knowledge is more cursory. It’s the career people who really run things. Most of them are solid Democrat, BTW. This goes up the line to the top career official at my agency who has made it known that he doesn’t like it when we go brief D.C. on our data and they grin and don’t ask the tough questions when we give them good news but take a totally different tack when it’s “bad” news.

    So I think the idea of a grand conspiracy driven by partisan politics is unlikely. If there is poor methodology on the surveys it’s more likely to be incompetence and poor professional oversight. I can’t speak for BLS, but where I stand, I just don’t see that happening.

  17. Mike commented on Nov 15

    The methodology of sampling one day of the month is perfectly fine for predicting a longterm trend. If the month of October doesn’t fully account for the increases you are hoping to see, November will more than account for it if the trend continues – it works both ways.

    From a statistics viewpoint, having more number crunchers collect 30 times more data will not improve the accuracy of the longterm trend that’s being tracked. Instead, it will just add more overhead with little results.

    All this crying is ridiculous…..

  18. Pat Gorup commented on Nov 15

    Down, huh? Regarding inflation, I was more interested in the FED’s infusion of another $47B today. It was the largest sum since 9-18-01. Does anyone have a running total on what the FED has “infused” this year? Or collectively; all central banks? So much attention is given to the FED interest rate game that I wonder; that of the two actions which has the best potential for being the most inflationary.

  19. Kent commented on Nov 15

    Pat Gorup, they “infuse” $47 bio and then take it back within a few days. It’s a repo.

  20. a guy called john commented on Nov 15

    i did an informal poll of businesses around my neighborhood. 8 out of 11 said they stock up on a full month’s worth of gas the Tuesday of the week where the 13th falls. the other three either it bought on the monday or wednesday of the same week!

  21. Winston Munn commented on Nov 15

    Pat Group,

    One more time, just to be helpful, the story of repurchase agreements – or repos.

    If you listen only to the headlines, you will be thrown off. First, these “infusions” are only short term loans. Second, to know the entire amount “infused” on a given day, the repos must be subtracted from the repo expirations of the same day plus any needed treasury demand.

    This “repo action” is not money thrown into the general economy – this is money lent to banks. The banks usually lend to each other, but when money gets tight or the collateral is suspect, banks are unwilling to lend or want to charge more than the federal funds rate.

    In that case, the fed steps in and makes repurchase agreements with the cash-strapped bank, taking collateral against a loan – usually an overnight loan, although 7-day and 14-day repos are also used.

    Today’s $47.25 repo action was offset by repo expirations (money that had to be paid back today) of $40.5 billion plus today’s needed treasury department demand, which was $6.5 billion.

    Net, there was only $0.25 billion of new money infused into the system today that could have gone toward stock or bond purchases.

    What most people don’t grasp is the significance of the huge repo auction today wasn’t that net money was infused into the system but that so much was needed to cover the banks’ inability or unwillingness to lend to each other. The significance is that it is getting harder and harder for banks to borrow the money they need to function.

  22. DavidB commented on Nov 16

    Why don’t they average over the month? Any reasonably bright 12-year-old could have figured that out.

    It is because we don’t have any bright twelve year olds in government.

  23. Jay in Cairo commented on Nov 18

    Part of this article is strikingly similar to another I read…that predates yours. Are you citing all of your sources?

  24. Anonymous commented on Nov 21


    Il 14 novembre la FEDERAL_RESERVE ha immesso nel sistema ben 47 mld di dollari con operazioni a brevissimo termine, accettando collaterali mortgagebacked per circa 23 mld di dollari. Questo dimostra ampiamente che la crisi solo all’inizio e la

Posted Under