Quality companies at bargain prices

I have just finished reading Joel Greenblatt's, "The Little Book that Beats the Market", and I must admit that I am annoyed I did not read this book earlier. It is certainly one of the best investment books I have read so far.

The central idea of the book is so simple, yet so true when you think about it. Greenblatt reckons that the key to outperforming the market over time is to buy businesses that are both high quality and selling at cheap prices. The methodology is equally as simple: screen the market for earnings yield and return on capital employed, rank every company on each measure and then add the ranks together. Companies with a low score on average outperform companies with a high score. It is so simple, but also so obvious. What I found most surprising (and to be honest I'm not sure I entirely believe it) is the extent to which this approach has actually outpeformed the market. Based on a back-testing exercise, Greenblatt claims that if you bought at the beginning of every year 30 companies in the top 1-2 deciles, held them for a year and then repeated the process again, you would have outperformed by 10-15 percentage points per year over the past 15 year period. What is even more surprising is that it works on the most basic measures - i.e. without any cyclical adjustment.

Importantly, he does admit that the approach does not work all the time. Indeed sometimes it underperforms for 2-3 year periods. Paradoxically, this could be the reason why it works over the long run, as most investors cannot handle such periods of underperformance and will give up.

For my part, I intend to use this approach for idea generation. It will be interesting to screen the market on these measures and see what ranks up the top (and down the bottom). I have just finished doing it for the US market, where I used 7-year average return on capital employed (ROCE) and 7-year median cash flow return on investment (CFROI) for the quality measure, and 7-year average earnings yield and last years EV/EBITDA for the valuation measure. Interestingly, when I compared the results with Greenblatt's own portfolio (as reported on 30/09/08) there was a pretty good correlation between my screen and his portfolio, suggesting that he may be using similar metrics. I found nearly 50% of his holdings in the top 3 deciles of my screen, versus only 25% in the bottom 4 deciles.