So. Incredibly. Cool. What if you could distill the strategies of the most legendary investment minds from the past century and then use build models to estimate how well they would have performed over the past few years?
The result would be a tournament between Warren Buffett, Ben Graham, Martin Zweig, Joel Greenblatt, David Dreman, Peter Lynch, Ken Fisher, James O’Shaugnessy, John Neff, and Joseph Piotroski.
This is the idea behind The Guru Investor, which is turning out to be one of my favorite books.
Professional investing isn’t about getting occasional flashes of insight, or relying upon the opinions of well-connected people. It is about having a statistically reasonable discipline and following it.
Most managers develop a set of screens that they use to find companies that are suitable for investment. One of the conclusions of this book is that there is no single statistic that can be used to find great stocks. Legendary investment managers typically seek a combination of factors, looking for attractively priced stocks (value) that are well-run (quality). Some, but not all, of these managers also look for growth.
The authors of this book researched what each of these managers have written about their own strategies, and then build usable screens to guess what they would have held in their portfolios during recent years.
As an example, John Neff ran the Windsor fund from 1964 through 1995. He generated annual returns of 13.7%, beating the S&P 500 by a over 3% per year for over three decades.
Here are the factors that Neff used in stock selection. (All of these screens needed to be passed in order to qualify for consideration):
1) Price/Earnings ratio >40% of the market average and <60% of market average (value)
2) Earnings Per Share growth >7% and <20% (growth)
3) Consensus forecast EPS growth >6% (growth)
4) Sales Growth > 70% of EPS growth rate
Sales Growth >7% (growth)
5) (Stock EPS growth rate + Dividend Yield) > ((Market EPS growth rate + Dividend Yield)*2)
(Stock EPS growth rate + Dividend Yield) > ((Industry EPS growth rate + Dividend Yield)*2) (value)
6) Positive Free Cash Flow (quality)
7) Positive year-over-year EPS growth over each of the past four quarters (quality)
All the bases are covered here – value, quality, and growth. How well might John Neff have performed if he was still in the business? According to this model, he would have generated average annual returns of 6.1% versus 3.2% for the S&P 500 during the 2004 through mid-2008 time period. Not too shabby!
Most investment books offer extensive analysis of how to buy stocks and much more limited analysis of what to sell. The Guru Investor is no exception – a single chapter is dedicated to sell discipline, which is defined as selling stocks that no longer meeting the selection criteria for purchase. In other words, if you wouldn’t buy it today, you probably shouldn’t own it.
So, how would these investment greats have performed over the past few years? According to the Guru Investor, a 10-stock model portfolio based on Ken Fisher’s methodologies would have generated annual returns of 20.8% per year from 2003-2008, narrowly edging out Ben Graham and beating the S&P 500 by a generous 16.8% annually.