Message Font: Serif | Sans-Serif
No. of Recommendations: 3
Holy smokes! I have come to realize that in another few months it will have been 10 years since I started with the Magic Formula Investment strategy. I have found the results to be very satisfactory with my personal investments.

On the other hand, the tracking portfolios using the alternate method suggested in the original book (ranking by ROA and P/E) have been a total bust. None of them worked out and there have been several up and down markets for the oldest portfolios, and both rising and falling markets for the more recent portfolios. Why did they perform so poorly? I don’t know, but maybe it is related to the market caps used, the time of year the portfolios were started, not enough stocks in each basket, bad input data, poor sorting by me, etc. There are probably many more reasons that might have been the cause. So, it is probably time to move on and call that experiment a failure. I will continue to track and report the performance through their anniversary dates, but I do not plan to continue maintaining them. Instead, I plan to start some new portfolios for tracking purposes.

My current MFI selections are one of the bright spots in my overall portfolio this year, having the strongest relative performance (losing the least) versus the indexes in the current YTD. The returns of my portfolio versus those of several Russell indexes through 30-Sep-2015 are as follows:

Portfolio Inception Date IRR R3000 R2000
Real Money 02-Feb-2006 +10.42% +6.56 +5.68

The following table details performance (IRR) over various time frames through 30-Sep-2015:
Portfolio 7 year 5 year 3 year 1 year
Real Money +16.16 +17.81 +24.96 +3.00
R3000 +9.91 +13.28 +12.53 -0.50
R2000 +8.63 +11.73 +11.02 +1.25

I also have several tracking portfolios that I maintain on Marketocracy as additional 'data points'. These were started on various dates using free data sources available from the internet. These portfolios are rebalanced annually near the date of inception with a small component left as cash to 'pay' expenses. Only portfolios with at least one year of history are included. The annualized returns of these portfolios versus those of several Russell indexes through 30-Sep-2015 are as follows:

Portfolio Inception Date IRR R3000 R2000
MFI 20-Mar-2006 +6.3 +6.42 +5.59
Fidelity 20-Mar-2006 +1.48 +6.42 +5.59
Yahoo 21-Mar-2006 +2.45 +6.44 +5.59
FinViz 30-Mar-2011 -2.29 +10.72 +7.95
MG1 08-Jan-2012 +11.41 +13.96 +12.37
SCHW 18-Mar-2012 -0.90 +11.50 +9.79

It seems like it is a constant race to find stock screeners that remain ‘free’. While playing around with FinViz I notice that they now require Elite access ($299.5/yr) in order to export screener results. That data point will fall by the wayside to reduce the work input required. It might have been like that for some time and I’ve just been cutting and pasting all the results into a text file and then importing into Excel. Though, looking over the results from the past that isn’t a big loss. At this point I’d have to say that the alternate ROA + PE screen isn’t robust and provides poor returns. This is the case for all current portfolios using those metrics: Fidelity, Yahoo, FinViz, SCHW.



R3000 and R2000 are the internal rates of return for the Russell3000® and the Russell 2000® Indexes, respectively, over the given time period.
My self-managed account consists of four groups of 10 stocks each that were purchased and spaced out about 3 months apart. The picks are from the MFI website and selected from the top 50 and then run through a list randomizer. The first 10 on the list that I did not previously own are selected for the next cycle. They are rebalanced/shuffled/changed out on their respective anniversaries. I do not continue hold stocks that are still on the list or buy the same stock that is held in the other three cycles.

Marketocracy fund notes:
Only the MFI portfolio uses actual magic formula stocks picked from the website. All the other portfolios use the general screening option method outlined in the book to pick your own stocks using ROA and P/E ratios. The MG1 portfolio consists of the 30 highest ranked stocks from M.Gerda who posts about MFI on his blog. There were several funds that may have somewhat incorrect overall returns due to an error in Marketocracy's software. These portfolios held a position that was acquired for cash and did not reflect in the transactions in a timely fashion (or possibly correctly).
The free Yahoo screener that I used for many years no longer exists to generate new portfolios starting in 2015. As a result the Yahoo portfolio is now cobbled together using a list of profitable companies (P/E > 0) and market cap above 250M and then data is gathered using the SMF add-in from Yahoo. This is an effort to try and keep the data stream as pure as possible.
Print the post  


What was Your Dumbest Investment?
Share it with us -- and learn from others' stories of flubs.
When Life Gives You Lemons
We all have had hardships and made poor decisions. The important thing is how we respond and grow. Read the story of a Fool who started from nothing, and looks to gain everything.
Contact Us
Contact Customer Service and other Fool departments here.
Work for Fools?
Winner of the Washingtonian great places to work, and Glassdoor #1 Company to Work For 2015! Have access to all of TMF's online and email products for FREE, and be paid for your contributions to TMF! Click the link and start your Fool career.