For the past couple weeks, I’ve written to you about several fundamentally attractive stocks based on Joel Greenblatt’s Magic Formula Investing (MFI) strategy. But what exactly is Magic Formula Investing, how does it work, and – more importantly – does it work?

In this article, we’ll briefly go over the strategy’s philosophies, implementation, and past results. Anyone interested in going further should pick up a copy of Greenblatt’s book The Little Book That Beats the Market. It’s the “Bible” of MFI.

The philosophy behind the strategy is very simple and straightforward: buy great companies at cheap prices. That’s it. It does not focus on particular industry sectors, does not concern itself with how big or small a company is, completely ignores forward earnings estimates, and does not discriminate between dividend payers and non-payers.

As you’re probably thinkging, buying great companies at cheap prices isn’t much of a strategy. That’s why MFI sets up the rules that define what makes companies “great” and “cheap” – and it does that automatically in the form of a stock screen. More on that in a minute…

So how do we tell what is a great company?

By measuring its return on invested capital (ROIC). At its core, a business is simply an attempt to generate higher returns on capital than can be obtained from alternatives. According to Greenblatt’s theory, those that can make the most money from each dollar invested are the best businesses. There are several measures of ROIC. One is return on assets – how much profit a company makes on each dollar of assets. A second, better one is return on equity, which measures how much a company earns on the net assets owned by shareholders. The best one is return on capital, which measures just the capital used in running the business, discounting excess assets such as large cash balances and non-tangible assets like goodwill. MFI uses a modified version of ROIC that tries to filter out various accounting vagaries that differ from company to company, measuring them on equal footing.

The second question is: how do we tell if a stock is cheap?

MFI uses a statistic known as earnings yield, a variation on the well-known P/E ratio, which measures how cheap a stock is against trailing one-year profits. Traditional earnings yield is just the inverse of P/E (E/P), expressed as a percentage. For example, a P/E ratio of 15 translates into an earnings yield of 6.7% (1/15). Earnings yield is useful in that in can be directly compared against the stated yield on alternative investments like bonds or treasuries. A high earnings yield indicates a stock that is priced cheaply against past earnings.

Greenblatt makes a few adjustments to his version. For one, he uses earnings before interest and taxes (EBIT), or operating earnings, instead of net income. This filters out non-recurring costs like goodwill impairment, and non-operating costs like tax and interest payments, focusing solely on the profitability of the core business.

Second, instead of using market capitalization, Greenblatt uses enterprise value, which rewards companies with a lot of cash and punishes those with a lot of debt.

One important caveat is that these two statistics are really most useful for traditional businesses that are valued using ongoing earnings. MFI removes financial companies where value is mainly measured by net assets. Banks, insurance companies, and utilities are explicitly filtered out.

Now we can automate the process using a stock screener, a computer program that finds stocks based on the criteria that we set.

Creating the Magic Formula screen is straightforward. Calculate the earnings yield and return on capital figures for all U.S.-traded, non-financial stocks. Rank them from highest to lowest, in each category, assigning a score (so the highest earnings yield would receive a ‘1′, for example). Add the scores together to get a composite score. Then sort this composite score from lowest to highest. (This is a fair amount of work, but you can skip the hard parts by visiting the resources in the P.S. of this article.)

With the resulting list in hand, Greenblatt recommends buying 3-5 of these stocks every few months, and then turning them over every year (primarily for tax purposes). The strategy as described is fully mechanical, with no additional vetting.

The final question is: does it work? In a word: yes, and extremely well. According to The Little Book That Beats the Market, following the MFI strategy from 1988-2004 would have delivered an average 30.7% annual gain, vs. 12.4% for the S&P 500. Since the strategy went public in 2005, MFI has continued to substantially outperform the market.

That is Magic Formula Investing in a nutshell. It is logical (buy great companies at cheap prices), simple to understand (just two metrics), easy to implement (turn over 3-5 stocks a few times a year), and has delivered exceptional outperformance (more than double the market’s return annually).

Maybe 2010 is the year to anchor your portfolio to this winning strategy?

Sincerely,
Steve Alexander
MagicDiligence.com

January 13, 2010

Steve Alexander

Steve Alexander is the founder and editor of MagicDiligence, an Internet-based research and recommendation service focused on finding the best opportunities from Joel Greenblatt's Magic Formula Investing screens.  By fundamental research of growth potential, competitive position, and financial health, his picks have vastly outperformed the market at a success rate far exceeding that of most mutual funds.  Prior to this, he was a successful independent stock investor for over 10 years.

View articles by Steve Alexander