Torturing the Data: Investors take heed!

 Here at "A Dash" my main mission is finding the best sources and accepting a wide variety of viewpoints.  I write a weekly update giving a perspective in various time frames.  My recorded positions vary over time, but I always use the same indicators.  If there is a change, it is a gradual process –carefully explained and reviewed.  This does not happen often.

I try to set an example for an objective approach.

A Disturbing Trend in Honesty

In recent weeks the stock market has declined.  Journalists and pundits have rushed to look smart by retroactively explaining the market move, or by reminding readers that they were "early" on the housing crisis (i.e. permabears).

This grab for the headlines, eagerly embraced by many journalists, has led to a disturbing outcome:

There is a bull market in torturing data!

Every day there is another parrot who dismisses actual data in favor of anecdotes and explanations of why this time is different.

There is a simple reason for this.  The parrot lives in a world of slogans and so do you.  You do not understand economics or causal modeling, so you have a bulls eye on your chest.  Repetition is crucial.  Interpreting data is hard.  Headlines are easy.  You are a target.

Here are some crucial examples:

  • The yield curve — long history, many historical periods including lots of Fed control.  Never a recession at current levels.  But this time is supposedly different.
  • Jobless claims — emphasized by bearish pundits when it had a five-handle but ignored now.  Never a recession at this level.  But this time is supposedly different.
  • Money Supply — perhaps the longest and strongest relationship.  Go read Milton Friedman to see the correlation with nominal GDP.  We can argue whether the result will be real growth or inflation, but something is happening.  Meanwhile — this time is supposedly different.

My Viewpoint as a Methods Prof

I offer this as a professional in research methods — student, model developer, teacher, and consumer of models for 40 years.  Most of the people you see on TV or read about have not taken a single class in research methods.  They have NEVER developed or tested a model.  Their prominence in the media reflects marketing rather than skill.  It is a sad commentary about who is featured on TV.  When I see someone I do not recognize I look at the bio.  Often it says something like "frequently quoted."  Sheesh!

It is intellectually dishonest to start with a conclusion and keep changing your methods until the data fit.  There should be a most wanted list of people doing this right now, including some of the most frequently cited sources.  You can spot them if you look.

A Journalistic Challenge

Here is an idea.  What if journalists looked for the cheats who change their methods and then wrote hard-hitting articles that highlighted the shifts.  That would be tougher than trolling for quotes, but more helpful to readers.  Maybe someone could give a prize for journalism that actually went against the grain.

Just a thought….ivory tower, no doubt.

Investment Implication

For many weeks I have highlighted the growing discrepancy between fundamental indicators of earnings, valuation, risk, and specific opportunities,  including two winners highlighted here and here.

I know from conversations and email that many investors are paralyzed with fear, unsure of how to get started.

I recommend that you formulate a plan that has the right size for you — one that can accept the expected volatility.  The next step is to find some stocks to buy.

Doing Nothing?

Do you expect inflation?  I do.  What rate are you getting on cash?  Most investors (and I say this only in general terms, since everyone is different) would benefit from a few great stocks with good dividends and good balance sheets.

For new clients I am still buying the stocks referenced — JP Morgan Chase (JPM) and Apple Computers, Inc (AAPL).  Despite my forecast last weekend, Felix let the gold miners (GDX) out of the penalty box, so we added to positions there.

On a strict valuation basis, even with a reduction in earnings expectations, I have not seen so many great opportunities since the bottom in 2009.

You may also like


  • jdb August 29, 2011  

    I think this is very sad. It’s tempting to think of it as being just a lack of character in the players. But, I wonder if it’s postmodern thought and those raised within it. It could produce a mental state where the concept of trying to find external reality is foreign to the mind.
    I read awhile back about researchers trying to determine the degree to which young people believed in the concept of external objective reality. But, they were surprised in the initial going that they had to spend time explaining the actual concept of that to the students. They couldn’t understand the concept itself. It wasn’t only that they didn’t believe in it, they had trouble actually conceptualizing it.
    Maybe, that’s a part of what is going on. If so, I think we are in a world of hurt.

  • louis August 30, 2011  

    Jeff, great post today!
    FWIW,I routinely refer to the “parrots” you describe as “one of the bobble heads on one of the OFTAA channels”.
    For example,”the other day I heard one of the bobble heads on one of the OFTAA channels saying XXX but a week before the same bobble head on another OFTAA channel said YYY”. I no longer refer to any of these guys by name nor name their platform.
    OFTAA channel? -ONE F#%KING THING AFTER ANOTHER channel.

  • rs August 30, 2011  

    Hear hear.
    Another reason to like Prof Jeff besides the intellectual honesty is his resemblance to the affable Wayne Knight (“Newman” but a moral, more handsome version).

  • Frank Newitt August 30, 2011  

    Great post. I recently read an interesting piece on the blog of an Australian boutique hf (Helix Partners) on the same subject that you might want to take a look at:

  • Doubtful August 31, 2011  

    Yes, the yield curve is “different this time”. How can you claim it’s not?
    It’s more or less impossible for the yield curve to invert when rates are at the lower bound. That’s why it’s different. We now have to look at Japan as the better model, not our own recent history. In Japan’s case, they went in to recession when the yield curve flattened.
    In many places where the yeild curve CAN still invert, it HAS e.g. many emerging markets. What does that tell you about the US?
    Also, on unemployment claims, I don’t see why you say “Never a recession at this level”:
    None of the above means that there HAS to be a recession, but I suspect you’re torturing the data just as much as the “parrots”.

  • oldprof September 16, 2011  

    Frank — Sorry to be so slow in responding to your comment. It has been pretty hectic. The link is excellent, and I have made a note for future reference.

  • oldprof September 16, 2011  

    Doubtful — Many observers who share your concern are looking at the slope in different parts of the curve. Try the ten-year/thirty year spread, for example. There are many, many indicators that have NEVER looked as they do going into a recession — corporate profits, balance sheets. I don’t like calling things 0 or 100%, but there are too many people taking a few data points and drawing exaggerated conclusions.
    As to the chart you cite, it is difficult to discuss in this format, but you obviously see something different from what most would. If you want to go back so far, you need some adjustment for the size of the labor force. We are in a bad place on job losses, but not the stuff of recessions.
    I try to let the data speak to me — and I do so every week, in objective terms, with regularly reported position changes.
    Thanks for joining in.