Magic Formula Investing

Joel Greenblatt's Magic Formula Investing has gained popularity since the release of The little book that beats the Market. It is a simple yet effective stock picking strategy combining investment philosophies from Benjamin Graham and Warren Buffett.

Let's analyze Greenblatt's strategy with the Stock Screening process so that you can build your own magic formula.

Investing Universe

Market Capitalization
Greenblatt backtested the strategy with the 3,500 largest companies, the smallest market cap was 50M. Greenblatt basically eliminated micro-caps.

The strategy works as well if you focus on Large Cap so use an Investing Universe adapted to your risk profile. Stick to Large Caps if you are risk adverse and include Small/Mid Caps if you can stomach higher volatility.

Primary Theme Go back to Top

Magic Formula is a Growth At Reasonable Price strategy using 2 parameters to pick stocks: Return on Invested Capital (Growth) and Earning Yield (Value).

Return on Invested Capital ROC is defined by Greenblatt as (Earnings Before Interest and Tax) / (Net Working Capital + Net Fixed Assets).

You may use Return on Assets ROA or Return on Equity ROE instead since they are more readily available at stock screeners. Greenblatt set the min threshold at 25% for ROA (ROC and ROE are often >ROA).

Caution: your screen may be skewed toward companies in industries that do not require much capital. You may want to put a cap on the max Return on Capital such as 50% or 100% (although Greenblatt does pick companies with ROC>100%).
Earning Yield is defined as (Earnings Before Interest and Tax) / (Market Cap + Debt). The denominator is also known as Enterprise Value.

You may simply use Price/Earning PE instead. Make sure PE is >0.

This interesting Magic Formula Investing Study by Robert Haugen shows that the strategy works as well if you use Return on Assets or Return on Equity instead of Return on Invested Capital and Price/Sales, Price/Cash-Flow or Price/Book value instead of Earning Yield.

Here, Haugen provides evidence that for a mechanical stock picking strategy to work, you do not have to religiously follow every criterion to the point but just build strategies that are reasonably similar to studies/backtesting.

Note: a popular parameter used by GARP investors is the PEG or (Price/Earning)/Growth. But since Greenblatt and Haugen found that ROC, ROA or ROE are excellent when combined with PE, we may be tempted to use PE/ROC or PE/ROA or PE/ROE in a screen.

The only difficulty would be to find the right threshold. In this case, it may be best to use these parameters in Ranking, which leads us to...

Ranking Go back to Top

Greenblatt actually does not use any filtering criterion in his strategy. Instead, he exclusively uses Ranking. Stocks are ranked based on a combination of High Return On Capital and High Earning Yield.

Now, very few stock screeners allow to rank a whole database based on a combination of factors. Most allow ranking with only one criterion.

A possibility is to export your screening results and perform Greenblatt's double Ranking in a spreadsheet application (e.g. Excel).

Alternatively, a solution is to rank with the PE/ROC that you can define in the field name of some screeners. Example with MSN Deluxe stock screener:

In this case, it is advisable to set a minimum PE (say PE≥5) or you may end up with companies with PE close to 0, whatever their ROC.

The beauty of ranking with the PE/ROC is that in one single criterion, you get both low PE and high ROC.

Finally, Greenblatt's alternative is to simply rank by low PE. And remember, the formula works if you use other Valuation measures.

Holding Period Go back to Top

GARP Investing, and therefore the Magic Formula, is adapted to holding periods ≥1 year. The inclusion of Value provides the necessary protection required over longer periods.