Stock forum

The preferences here were, from my understanding, Graham's personal preferences based on his own research. Number 5 has been researched (alone) extensively. Also 1, 3 and 6 (and 9) have also been researched and have shown to be good results. If Lowe has done any other studies I'd be interested in the results.

And if you happen to look at Jae's recent article, he does include 1, 3, 6 and 7 in one of his screens at least during the 6 year period he studied, it outperformed 1, 3 and 6 (though I doubt the difference is statistically significant.) It would be interesting to look at it over a wider period.

And I'm a big fan of testing ideas and critical thinking and the like so I like to see ideas testing, I like seeing evidence, etc. That's why I've emphasized the criteria I have. I'm always willing to look at more evidence.

And just to be thorough (another one from JSTOR unfortunately):

Ben Graham's Net Current Asset Values: A Performance Update

This looks at #5 which everyone here should already be familiar with since Jae has a similiar screen. Alphas are calculated and are generally positive.

The only interesting point here comes from James Montier's research. One of the claims the efficient market guys have made is that none of this indicates that markets aren't efficient but instead that there is some other element to risk other than Beta. Montier, himself, is quite critical of using Beta as a measure of risk and prefers Graham's definition of risk as a "permanent loss of capital". Montier gave the following operational definition of this (for the purpose of his study) as being a 90% or more loss in one year.

Montier found that for the market as a whole, only 2% of stocks suffer a permanent loss of capital (in the sense defined above) whereas 5% of Graham's NCAV's suffer a permanent loss of capital. So they could very well be considered "riskier" in that sense. Granted, in spite of that fact, the portfolio performed better than the market.

Here's a link to the screen (hopefully this works): Graham 1-3-6-mod

If that didn't work, here are the stocks that made the screen:

ASR
BOBE
ELP
GRMN
INTC
MAIL
WMK

By adding Altman and Piotroski filters, I reduced a list of almost 100 stocks to a list of 7 stocks, for better or worse.

With regard to the article at JSTOR, if you give me an email address I can send you a copy (being a student has some benefits). But in all honesty, there isn't too much different. The study looks at 1972-1981 for NYSE-AMEX as opposed to 1977 to 1995 for the JSE in the other article. The results are similar (market beating returns, positive alphas, etc).

The only interesting addition is that they attempt to add criteria 9 to 1, 3 and 6 and it made the results worse. So 1,3, and 6 is better (at least for that time frame.)

Montier did his own study which he discuess in his value investing book with 1, 3 and 6 and got similar results (for a different time period). He also found that adding the criteria of 10 year P/E ratio of less than 16 improved the 1, 3, and 6 criteria.

I'm not sure if adding Altman/Piotroski will be beneficial or not but it does make it a more manageable list to make selections.

Post edited 3:09 pm – August 31, 2010 by Jae Jun

Apart from #5 (which is the NCAV criteria), Graham favored 1, 3 and 6. Here are two papers showing the effectiveness of these criteria in stock selection. Alpha and beta are calculated to measure risk adjusted returns (granted, in the words of James Montier, "CAPM is CRAP".)

A Test of Benjamin Graham's Stock Selection Criteria (1984)

A test of Graham's stock selection criteria on industrial shares traded on the JSE (1997)

James Montier found that results to criteria 1,3 and 6 could be improved by requiring issues to have 10 year P/E of less than 16 (another one of Graham's criteria.)

On a sidenote, on the Robot Dough Screener, I have a screen set up that looks at 1, 3 and 6 along with Altman and Piotroski for financial strength. I used 3% for AAA bond yields.

Jae; The following is taken verbatim from the book "Value Investing Made Easy"  p.100-101 by Janet Lowe who is an avid follower of Ben Graham.  A scoring system could be created as a screen or this score could be created in the "Graham Valuation" or "NetNet" spreadsheets. Here we go:

Ten Attributes of an Undervalued Stock

Professional investors have their favorite places to look for extra protection. An investor may accept different sources for that margin, depending on the circumstances and the investor's own zone of comfort. High asset value, strong cashflow, and dominant market position are among the influencing factors.

The first step is to examine the quantitiative factors we've discussed earlier (net asset value, working capital, proce-to-earnings ratio, debt-to-equity ratio, and so on). graham made the following list of attributes of an undervalued stock. He noted that any company that meets 7 out of the 10 criteria is undervalued and has an adequate margin of safety.

Criteria 1 through 5 measure risk, 6 and 7 establish financial soundness, 8 through 10 show a history of stable earnings. Very few companies meet all 10 criteria.

1. An earnings-to-price yield (reverse of P/E ratio) that is double the AAA bond yield. If the AAA bond yield is 6 percent, the earnings yield should be 12 percent.

2. A price-to-earnings ratio that is four-tenths of the highest average P/E ratio achieved by the shares in the most recent 5 years.

3. A dividend yield of two-thirds the AAA bond yield. Stocks that lack either a dividend or current profits are automatically eliminated by this rule.

4. A stock price of two-thirds the tangible book value per share. The formula for tangible book value is [Total Assets - Intangible Assets - All Liabilities - Stock Issues]/[# of Common Shares Outstanding].

5. A stock price that is two-thirds the net current asset value or the net quick liquidation value. This was the earliest investment technique used by Graham.

6. Total debt that is lower than tangible book value.

7. A current ratio of two or more. This is a measure of liquidity, or a company's ability to pay its debts from income.

8. Total debt of no more than the net quick liquidation value.

9. Earnings that have doubled in the most recent 10 years.

10. Earnings that have declined no more than 5 percent in 2 of the past 10 years.