Backtesting Ben Graham
Charles Mizrahi, over at http://www.hiddenvaluesalert.com/, suggested we backtest a simple Ben Graham strategy mentioned in a 1976 article he dug up in Medical Economics. Charles has been implementing Ben Graham related strategies for many years and the live performance of his recommendations are monitored by the Hulbert Financial Digest (HFD), which helps investors sift between the “good”, “bad”, and “ugly” of the newsletter world. His newsletter, “Inevitable Wealth Portfolio” (IWP) is top-notch and worth a look for dedicated deep value investors. According to HFD, IWP is up +107.4 versus the +S&P 500 73.1% from Feb. 1, 2009 through Apr. 30, 2011.
Here is a link to the modern day version of the Medical Economics magazine where Ben Graham mentions how to implement his simple strategy.
So what’s Graham’s secret to achieve 15%+ returns over long horizons?
Well, below is an excerpt from the original 1976 article with all important points highlighted:
[gview file="http://turnkeyanalyst.com/wp-content/uploads/2011/04/Simple-and-Easy-Approach-Medical-Economics-Graham-1976.pdf"]
We decided to keep it simple and backtest the low P/E (<10), shareholder equity > .5 strategy from 1965–2010. We also backtested the results in accordance with the “trading rules” alluded to by Graham: stocks entering the portfolio are held for 2 years, or if they appreciate >50%. For robustness, we tested a variety of P/E and shareholder equity combinations–all results are very similar.
Here are some highlights from the analysis:
[gview file="http://turnkeyanalyst.com/wp-content/uploads/2011/04/postdata1.pdf"]
We plan to write up an academic article this summer that goes into the details of our final results and analysis. We wanted to share the “hot off the press” results with readers of the Empirical Finance Blog™.
Enjoy!
test
This is neat. A lot of people use Graham’s method, or some variation (eg Buffett), and it seems to work but I’ve never seen the methodology backtested before.
This reminded me of a recent article on Joel Greenblatt, the “modern day Graham”, and his release of 4 new mutual funds which pick stocks based on his Magic Formula. Compared to Graham, it looks like Greenblatt looks at profitability (return on capital) instead of financial soundness (as measured by Equity/Assets), and uses earnings yield instead of P/E as a measure of value. And apparently it has a 4.5% alpha when calculated by the FF 3-factor model… interesting
Hey,
I created a screen to show current Ben Graham formula stocks. You can see it here- http://www.whopperinvestments.com/ben-graham-formula-stocks#more-503
Hope it helps. Keep up the create work on the blog!
Nice work.
If you go looking for it, you will find a slightly later set of criteria advocated by Mr Graham. The criteria are sometimes referred to as the “last will and testament” and can be found in this article:
http://www.scribd.com/doc/33538620/Value-Avatar
The trick here is to produce a list of stocks which satisfy any one of criteria 1-5 AND and one of criteria 6-10.
The “last will and testament” is also referred to in J Train, “The Midas Touch” (Perennial Library, 1987) at pp 11-14, which also refers to some academic back-testing work.
Cheers.
Where can we see the results? The link seems to be broken.
Thanks!
I enjoy your blog and find that the results are quite interesting indeed. I wanted to find out whether or not you will be posting the academic write-up. Another interesting possibility would be to run Greenblatt’s MF and the PV strategy for the same time frames as a comparison. Thanks again for making this backtest available!