Sunday, 27 September 2015

Greenblatt Magic Formula Experiment, Sep 25, 2015

I'm going to regress somewhat by doing a post on an attempt at building an auto-pilot portfolio courtesty of "The Little Book that Still Beats the Market"

I'll call this a real time experiment with imaginary money.

I've done a post on Joel Greenblatt before.  He was interviewed in Barrons last summer, see link here.

I'm going to follow Greenblatt's system to a tee on paper, which is basically the equivalent of throwing darts at 20 stocks with combined characteristics of highest return on capital and highest earnings yield relative to every other stock in Greenblatt's ranked universe.

I am in no way, shape or form, advocating that anyone throw darts with real money or copy this portfolio.  This is purely an educational exercise.  The system works as follows:

  • Run screen of 50 highest ranked Greenblatt universe stocks at
  • Buy top 20 of the 50 highest ranked stocks on day 1 of the year.  As we're now basically at Oct 1, 2015, I'll make Friday's close, Sep 25, 2015, my day 1 of the year.
  • Hold top 20 stocks for one year
  • On Sep 24, 2016, re-run the screen.  Turn the existing portfolio over and book the gains/losses, and buy the top 20 results from the new screen
A few comments regarding the results, which I've reproduced here:

  • I cheated by purposely ignoring a few of the stocks which I believed to be absolutely hopeless, such as ITT Education, Capella Education, Collectors Universe, and Lifevantage.  It will be interesting to see whether excluding these stocks ends up detracting from the paper performance of the portfolio
  • In their place, I simply selected the next lowest ranked stock 
Here's the portfolio on day one, assuming I bought everything at the close on September 25, 2015.  I will try and figure out how to make this portfolio accessible by seperate tab on the blog (I'm still learning how to use blogger).

Why am I doing this at all?

Curiosity mostly.  I currently have a number of personal accounts under my discretion, and it seems that the accounts I don't tinker with (the shareowner accounts), do better than the accounts I actively tinker with.  As I've developed an appreciation for Greenblatt, I figured I'd run his system on paper to see whether or not "The Little Book that Beats the Market System" actually  beats the market in real time.


  1. I have tried to see about using the MF through some testing and have never had good results. If you check out you can see their version, which at initial glance looks good but over time has not seemed to stand up. In addition, no one has been able to replicate the results that Greenblatt presented in the book.

    The other problem I have had, which you alluded to, is that sometime the stocks that get selected are total junk and or even over-the-counter which some accounts don't allow you to hold.

    Nonetheless, I am going to follow with great interest your results here.

    1. I also wonder if any advantage hasn't already, by virtue of being an advantage, been arbitraged away over time? i.e., Greenblatt publishes book, demonstrates alpha, systems bid up companies shown to have Greenblatt type characteristics over time. Fast forward to 2015, system doesn't work...

  2. Thanks Jeremy, it's basically a 360 degree turn and in complete contrast to how I feel my approach has evolved over time, and I'm doing this moreso to see how it holds up under market duress. When I first read the little book last summer, I thought wow, what a great sounding system. Now, I'd be pleasantly surprised if all of the 20 stocks above end up at par come next September.

    Thanks for comment and keep in touch.