The non-magic formula approach

There is no magic in investing…but there is a common sense way to filter the universe and build a portfolio that seems (if it works) magical. Inspired by Joel Greenblatt’s ‘magic formula’ presented in The Little Book that Beats the Market, I’m constructing a concentrated portfolio of public equities selected using simple, quantitative criteria. Mainly. I’m still fine-tuning my own ‘magic formula’ so, for now, I reserve the right to tinker (likely to my detriment, but that’s where I am at the moment).

In his book, Greenblatt sets out to find a purely objective way to find good companies at bargain prices. That seems to me a reasonable approach. And his results, at least at time of publishing, seem to legitimize the approach.

Greenblatt’s magic formula screens for companies that are capital efficient (high return on capital) and selling at a reasonable price (low Enterprise Value to Operating Income ratio). He then ranks them accordingly, aggregates their rankings, and buys from the top.

He recommends a basket of at least 20 securities that are rotated in and out on an annual basis.

Others have followed, attempting to improve on Greenblatt’s simple approach by substituting his two-variable formula with a more sophisticated (i.e., complicated) method of objectively filtering the universe and homing in on the best opportunities (see Quantitative Value, Gray & Carlisle). Their logic is sound and I’m sure their math is too. The risk is overcomplicating things.

I’ve tried to incorporate some of the ‘safety-first’ filters that Gray & Carlisle propose, like scrubbing businesses whose financial statements show potential signs of manipulation, or those showing real signs of financial distress, but getting the data is a challenge. Still a work in progress.

How it works, in theory:

1. Once a quarter, filter the universe using simple, objective criteria. Example criteria might be:
   1. Safety filter: e.g., Market Cap >$100M
   2. Quality filter: e.g., Minimum Return on Assets (ROA) of 25%
   3. Bargain filter: e.g., Earnings Yield >10%
   4. Exclude certain sectors or industries (e.g., Financials and Utilities)
2. Rank the results based on your Quality and Bargain filters.
3. Combine their scores (the lowest score any company can have is 2: Ranked 1st for Quality + Ranked 1st for Bargain)
4. Buy the Top 5
5. Sell the 5 companies you bought 12 months ago (note: There’s some fine tuning to do here if you’re running this in a taxable account. You’ll likely want to sell winners after a year has passed to lock in long-term gain tax treatment, and sell losers before a year has passed to lock in short-term taxable income offsets).
6. Rinse. Repeat.

Why it works

1. It focuses on quality businesses selling at reasonable prices.
2. It reduces the odds of making irrational snap decisions.
3. It is designed to take advantage of Mr. Market’s mood swings.
4. It’s tax efficient (if you pay attention).

Why it's difficult

1. Data: You have to find a source of data that makes it easy to filter the universe using high quality data. You’d be surprised what kind of challenges even the largest data providers wrestle with (at least I was). To give yourself the best chance, make sure your data source is either first hand (e.g., SEC filings) or from a reputable vendor. If you’re using Greenblatt’s free screen provided at (magicformulainvesting.com) then this is less of an issue although it’s more opaque.
2. It’s tempting to tinker with the formula endlessly. Tinkering is just as likely to harm your results as it is to improve them.
3. It’s tempting to cherry pick, to allow subjectivity to enter the equation rather than let the objectivity of the formula do the work.
4. You won’t know for a few years whether it’s actually delivering above-average returns over extended periods. Nothing works year in and year out, but it needs to work over rolling 5 and 10 year periods in my book.

Evidence this approach (probably) works:¹

1. Joel Greenblatt’s Magic Formula achieved a 19.7% compound annual growth rate (CAGR) from 1988 to 2009 compared to a 9.5% growth rate for the S&P 500. Better still, when he included companies with Market Caps less than $1B (over $50m), the annual rate of return was 23.8%. At that rate, you’d have 100 times your money in about 20 years.
2. Wesley Gray, PhD & Tobias Carlisle dubbed their model “Quantitative Value” and, according to their backtest, that model would have achieved a CAGR of 17.68% from 1974 to 2011, compared to the S&P 500 Total Return of 10.46%.