Magic Formula: An offshoot of Value Investing

Steve Forbes at Forbes.com has been conducting a number of fascinating interviews recently, with various famed value investors. In a recent interview, I came across a description of the "Magic Formula" investing strategy developed by Joel Greenblatt in his book "The Little Book that Beats the Market".

The magic formula reduces the value investing process to just 2 variables: Earnings Yield (Inverse of P/E) & Return on Capital.

A guy named Shankar had set up a website for Indian magic formula stocks called www.magicformulaindia.com. This website used to replicate the Magic Formula methodology and spit out a list of stocks which met the criteria laid out by Joel Greenblatt. Sadly, this website is not up and running anymore, as Shankar has moved on to other (presumably better) stuff.

A description of the Magic Formula way of Investing, straight from the horse's mouth follows (from the Forbes.com interview):

Greenblatt: This is what I wrote the book about five years ago. It follows the principles I've used in investing since I started my firm in 1985, and I've been teaching at Columbia for the past 14 years and these are the principles I've used to teach. About seven or eight years ago, we set out to test it. The two things I look at are earnings yield, which is how cheap is the company. A simple earnings yield would be the inverse of the P/E ratio or earnings to price. So, in other words, if something earned $2 and it cost you $10 a share, you'd have a 20% earnings yield.

We use a more sophisticated metric than just earnings, than just price. But the concept is the same. We use EBIT--earnings before interest and taxes--and we compare that to enterprise value, which is the market value of a company's stock plus the long-term debt that a company has. That adjusts for companies that have different ratios of leverage, different tax rates, all those things.

But the concept is still the same. We want to get more earnings for the price we're paying. That was sort of the principles that Benjamin Graham taught, meaning that cheap is good. If you buy cheap, you leave yourself a large margin of safety. Warren Buffett had a twist on that and said, "Gee, it's nice to buy cheap things but I also like to buy good businesses." So if I could buy good businesses at a cheap price, it's better than just cheap.

We have another metric that we look at in the magic formula which tries to determine those companies that are in better businesses. So we look at a business' return on capital. I described in my book the example that I gave to my son which is he had a friend that used to buy a pack of gum before he went to school with five sticks of gum in it. And he paid 25 cents for the pack of gum and at school he'd sell each stick of gum for 25 cents. So he would collect $1.25 and the pack would cost him a quarter and he would make a dollar for every pack he sold in school.

So we supposed that his friend grew up and opened a chain of gum stores. And let's say to set up a store with inventory in the whole store, displays and everything else it costs $400,000 to open a gum store. And each year one of those gum stores earn $200,000 a year. So the cost of setting up the gum store is $400,000 and each year it spits out $200,000 a year in profit. So that's a business that earns a 50% return on capital.

Then I said, "Let's suppose we have another business and I called that 'Just Broccoli' and it was also a store but only sold broccoli." But it still cost $400,000 to open a store. Each year, that store only earned $10,000. So that's a 2.5% return on capital. So we simply say that a business that earns the higher return on capital [is better].

So what the magic formula does is ranks companies first according to cheapness based on their earnings yield; the higher the earnings yield, the cheaper it is. So you rank all companies, thousands of companies based on how cheap they are based on their earnings yield. Then in a totally separate way you rank all companies again, but this time you're just ranking them based on their return on capital; the higher the better. So you do two rankings--one just based on cheapness and one based on return on capital. Then what the magic formula does is combine those rankings. So, in other words, if you were the cheapest company on this list, you'd get a 1 for cheapness and if you were the 200th highest return on capital, your combined score would be 201. If you were the 60th cheapest out of 2,000 companies and you were the 60th best return on capital, you'd get a combined score of 120. So you want to get close to 2, which is the best you could do. So the 120 score would come out better than the company that was the cheapest in the universe, which was 201.