How Do The Magic Formula Investing Rankings Work?
After 6 years of operation and hundreds of articles, I recently became aware that MagicDiligence had never posted an article detailing just how the strategy we followed, Magic Formula® Investing (MFI), actually works! It is time to remedy that today.
First of all, some background. MFI was invented by very successful former hedge fund manager and current professor Joel Greenblatt. He outlined the theory and method to the strategy in his 2005 book The Little Book That Beats the Market, which was subsequently updated in 2010. I highly, highly recommend anyone with even a passing interest in the strategy to read the book. It is quite short, under 200 pages, and takes all of 2-3 hours to read. Greenblatt is an entertaining and straightforward writer. It is one of the best investment books out there, for anyone.
Greenblatt found that when he did a ranking of the stocks in the market based on two simple statistics - EBIT/EV earnings yield and pre-tax return on tangible capital - the stocks at the top of the composite rankings annihilated the S&P 500's returns. This was no small difference. MFI stocks generated a 30% annual return from the study's 1988-2004 trial period, vs. the S&P's 12% figure. What's more, when breaking the ranked stocks down into deciles, there was a perfectly linear correlation between ranking and performance.
The "official" stock screen for MFI is still available for free here. From it you can generate any list of stocks from a minimum market cap, showing you the top 30 or 50 stocks in the market ranked by MFI statistics.
I think the ranking system is easiest to explain with an example. Instead of most of the entire universe of U.S.-listed stocks, let's do it with a very limited sample: the "big 5" defense contractors, namely Northrop Grumman (NOC), Lockheed Martin (LMT), General Dynamics (GD), Raytheon (RTN), and Boeing (BA). This example was done using the Portfolio Rankertool.
The first thing to do is to calculate and rank each of these stocks by their earnings yield (EY, see this article for the full details behind this calculation). Here are the results:
NOC 11.3%, rank 1
RTN 10.7%, rank 2
GD 9.2%, rank 3
LMT 9.0%, rank 4
BA 6.8%, rank 5
Now we do the same thing, separately, for pre-tax return on tangible capital (ROTC,explanation here). The rankings here are:
RTN 66.6%, rank 1
NOC 56.8%, rank 2
GD 46.9%, rank 3
LMT 30.3%, rank 4
BA 6.8%, rank 5
So, for the composite rankings, we simply add the rank score together for both the earnings yield and return on capital lists to get a composite, then order the stocks from lowest to highest. The final, MFI-style ranking looks like this:
RTN score 3 (2 EY + 1 ROTC)
NOC score 3 (1 EY + 2 ROTC)
GD score 6 (3 EY + 3 ROTC)
LMT score 8 (4 EY + 4 ROTC)
BA score 10 (5 EY + 5 ROTC)
Raytheon and Northrop actually tie for the best MFI ranking - this is not uncommon. Boeing scores as the least attractive of the "big 5" from a statistical standpoint. It is both the most expensive vs. earnings and the least efficient in terms of earning profits on invested capital.
Now, if you can imagine expanding our set of stocks out to all of the NYSE, NASDAQ, and AMEX listings, (minus financials, utilities, and foreign firms with ADR listings), then applying the same process, you understand how the MFI rankings work!