Graham's Last Will & Testament
Wayne Thorp will speak at the 2015 AAII Investor Conference this fall; go to www.aaii.com/conference for more details.
While researching Ben Graham’s net current asset value approach to investing for Computerized Investing’s February 2010 Online Exclusive (available online to Computerized Investing subscribers at www.computerizedinvesting.com), I ran across an article written by James Rea shortly after Graham’s death in The Journal of Portfolio Management.
Prior to meeting Graham a few years earlier, Rea had been working on a stock selection methodology that looked for companies with high reward-to-risk ratios. Upon reading an article that Graham had written in Barron’s—“Renaissance of Value”—Rea discovered that his approach seemed similar to Graham’s. On a lark, Rea forwarded his research to Graham. A couple of months later, Graham called Rea and asked how it was that he was finding his kinds of stocks and suggested that they meet. That first meeting led to a three-year working relationship, which culminated in Graham and Rea starting an investment fund that used the “best” stock selection criteria based on their research.
This article outlines the 10 criteria Graham and Rea first developed (and which they tested using a 50-year period), and builds a screen with the three criteria that they used for their Rea-Graham Fund.
Picking Stocks Like Picking a Milk Cow
In Rea’s article, he recounts the first time he met Ben Graham. Graham asked him to describe his stock selection theory, but without using mathematics. Rea did so using his “milk cow” analogy.
Rea likened his method of picking stocks to that of selecting a milk cow. When picking a milk cow, you ideally want a cow that gives lots of milk (earnings), but also will give more milk year over year (growth in earnings). Furthermore, you would like stability in the growth. Also, Rea went on to explain, cheese can be made from the milk, which is the dividend given back to the customer. So, he also would want a lot of cheese from the milk (a high dividend yield).
Part of Rea’s research was to look for stocks with high reward-to-risk ratios. To him, a “high reward” cow was one that produced a lot of milk, increased its milk output at a stable rate, and allowed for a lot of cheese to be made.
The risk in Rea’s milk cow analogy was that the cow would stop producing milk, meaning you lose your milk (earnings) and cheese (dividend yield). The degree of risk was how much you paid for the cow relative to what you could get for the “meat on its bones” (the liquidation value). The higher the price paid for the cow relative to its “liquidation value,” the higher the risk.
To complete his analysis, Rea divided his reward measure (earnings, growth, stability and dividend) by his risk measure (the price paid for a company relative to its liquidation value) to arrive at a relative measure of “how good” a company was relative to other companies.
Over the next several months after their initial meeting, Rea and Graham developed a set of 10 basic investment criteria they felt would be useful for selecting stocks.
Among these 10 criteria, five were susceptible to changes in earnings and price while five were not. The first five tend to change rapidly with changes in price and earnings.
The First Five
One criterion Rea and Graham used required that the earnings yield be at least twice the average AAA corporate bond yield. The earnings yield is the inverse of the price-earnings ratio, so that a price-earnings ratio of 20 would result in an earnings yield of 5% (1 ÷ 20 = 0.05, or 5%).
If the AAA bond yield is 5.5%, a stock would need to have an earnings yield of at least 11% to meet the Rea- Graham criterion. This translates into a price-earnings ratio of no more than 9.1 (1 ÷ 0.11 = 9.1).
The second criterion used the two-year average high price-earnings ratio, or the average of the high price for year 1 divided by year 1 earnings and the high price for year 2 divided by earnings for year 2. A stock should have a price-earnings ratio that is 60% below its previous two-year average high. In other words, the current price-earnings ratio had to be down to at least 40% of the previous two-year high.
As Rea explained it, Graham initially was only interested in price movement, looking for stocks that were down at least 50% from their previous high. However, Rea countered that it was illogical to require a company’s price be down 50% from its high when earnings were growing rapidly. Instead, Rea suggested requiring the price-earnings ratio to be 70% below its high. The two eventually compromised and used 60% below the average price-earnings ratio high.
Discount to Tangible Book Value
The third criterion called for the price to be less than or equal to two-thirds of the tangible book value.
Throughout Graham’s career, he was always interested in a margin of safety when buying stocks. Typically, this involved buying stocks trading below their “liquidation value.” When estimating the liquidation value, Graham focused on tangible assets—property, plant and equipment—the value of which could be reasonably estimated. Intangible assets such as goodwill, patents, trademarks, etc., are more difficult to value, which is why Graham focused on tangible assets.
The fourth criterion of the “first five” required the dividend yield to be greater than or equal to two-thirds of the average AAA bond yield. Therefore, if the AAA bond yield is 5%, a stock must have a dividend yield of at least 3.35% (5% × 0.67).
Net Current Asset Value
The fifth criterion Graham and Rea used called for the stock price to be below the company’s per share net current asset valueor “net quick” asset value. This is the “meat on the bones” a company would have left if it failed and is an approximation of what shareholders would be left with if the company went through a quick liquidation. They defined this as:
NCAV = Current assets – current liabilities – long-term debt – preferred equity
Assuming a company’s working capital (current assets less current liabilities) is conservatively stated, Graham and Rea felt that a firm could reasonably be expected to be sold off for the value of these assets. Furthermore, they also felt it was reasonable to expect that the company’s remaining long-term assets would fetch enough to offset “shrinkage” in current assets resulting from their conversion into cash.
The Second Five
The second set of five criteria Graham and Rea tested was not susceptible to changes in price and earnings.
The first three from this group measure a company’s financial strength or soundness. In order for Graham and Rea to consider a company “financially sound,” it had to meet at least two of these three criteria.
The first criterion from this second set sought out companies whose current ratio is greater than or equal to two. The current ratio is current assets divided by current liabilities.
Debt to Equity
The second of these five criteria called for a total debt (total liabilities) to be less than the company’s net worth. Or, in other words, the debt-to-equity ratio is less than or equal to one.
Debt to NCAV
Previously, Graham and Rea looked for companies with stock prices below their net current asset value. The third criterion of this second set of five required that total debt (total liabilities) be less than the net current asset value.
The final two criteria Graham and Rea tested dealt with earnings growth. The first required that companies had doubled their earnings over the last 10 years. This translates into an annualized growth rate of at least 7%.
The final criterion required no more than two declines in annual earnings over the last 10 years. Furthermore, they defined a decline as a year-over-year drop of 5% or more.
Graham and Rea knew it would be difficult to find companies that would satisfy all 10 of these criteria. As a result, they developed a scoring scale to identify potential investments.
When analyzing companies based on the 10 criteria, Graham and Rea awarded whole and half points. If a company met the criteria as they were presented earlier, they received a whole point. Rea states in his article that it is also possible to award a half point to a company, depending on how close a company is to meeting the exact criteria. However, Rea offers explicit examples of awarding half points for only two of the criteria—dividend yield and earnings growth.
In order to receive a whole point, the dividend yield must be at least two-thirds that of the AAA bond yield. However, a company could receive a half point if the dividend yield is between one-half and two-thirds of the AAA bond yield.
One of the earnings growth criteria states a company can have no more than two 5% or greater annual declines in earnings over the last 10 years. However, if it has three declines during this period, Graham and Rea would give it a half point.
Graham and Rea began their screening process by looking at the “second five” criteria. In order to be considered at all, stocks trading on the New York Stock Exchange had to score at least three and one-half out of five; a score of four or more was required for American Stock Exchange and over-the-counterstocks (the NASDAQ exchange did not begin trading until 1971).
If a company passed this first round of scoring, the “first five” criteria were looked at. It was considered a “buy” if it traded on the New York Stock Exchange and scored at least three and one-half out of five for the “first five” criteria. For American Stock Exchange and OTC stocks, it was a “buy” if it scored four or more on the “first five” criteria.
Testing the Factors
Graham and Rea intuitively knew that some of the 10 criteria were more important than others. To prove this, Graham suggested they test each criterion individually over a 50-year period from 1925 to 1975 to determine the relative importance of each. Here is a summary of what they found:
- Buying stocks with an earnings yield at least twice that of the AAA bond rate would have generated an average compound growth in price over the 50-year period of 19.9%, versus 7.5% for the Dow Jones industrial average;
- Buying stocks where the dividend yield was at least two-thirds the AAA bond yield would have generated an average compound growth rate of 19.5%; and
- Buying stocks with a price less than or equal to two-thirds of the tangible book value would have generated an average compounded growth rate of 14.2%.
A Simplified Approach
Graham and Rea knew their 10-point system was too restrictive and cumbersome for practical application. Therefore, after conducting their research of the individual criterion, they decided upon three criteria, which served as the cornerstone of the Rea-Graham Fund:
- Earnings yield of at least twice the AAA bond yield;
- Dividend yield of at least two-thirds of the AAA bond yield; and
- Debt-to-equity ratio of less than one.
Diversity Over In-Depth Analysis
Another reason why Graham and Rea limited their selection criteria to three instead of the original 10 was because it provided them with a larger pool of investment candidates. Graham was always an advocate of diversification, investing in as many as 100 companies or more if the market conditions were right. Rea said that Graham was always suggesting:
“Just buy a little bit of each, really not trusting it too much, because there may be many other reasons, that you don’t know about, as to why the company might not do as well in the future.”
Toward the end of his life, Graham had grown skeptical of the benefits of in-depth company analysis. In fact, he was quoted as saying:
“I do not have much confidence in the practical worth for most analysts of detailed studies of individual companies, with emphasis placed on either their comparative performance or on predictions of their relative future performance over a one-to-five-years time span.”
Instead, Graham felt that diversification was far more important than in-depth research. As a result, Rea and Graham always tried to invest in at least 30 companies at a time.
Building a Screen
Using AAII’s Stock Investor Pro fundamental stock screening and research database program, we recreated the Rea-Graham criteria. Using the five-year AAA corporate bond rate of 3%, the criteria for our Rea-Graham screen are:
- Earnings yield of greater than or equal to 6% (or price-earnings ratio less than or equal to 16.7);
- Dividend yield of greater than or equal to 2%; and
- Debt-to-equity ratio of less than one.
We excluded financial firms, REITs, OTC stocks, and foreign companies. We used the Yahoo! Finance Bonds Center (finance.yahoo.com/bonds) as our source for the five-year AAA corporate bond yield.
Profile of Passing Companies
Table 1 presents the characteristics of the stocks passing our Rea-Graham screen as of March 26, 2010.
In order to pass the Rea-Graham screen, a company must have an earnings yield of at least 6% (twice the current AAA corporate bond yield of 3%; alternatively, it must have a price-earnings ratio less than or equal to 16.7). The median earnings yield for the companies passing the screen as of March 26, 2010, is 9.0%, compared to an earnings yield of 5.3% for the typical exchange-listed stock. This translates into a median price-earnings ratio of 11.1 (100 ÷ 9) for the current passing companies.
|Portfolio Characteristics (Median)||
|Earnings yield (%)||9.0||5.3|
|Dividend yield (%)||3.2||0.0|
|Price-to-book-value ratio (X)||1.8||1.6|
|Price-to-sales ratio (X)||1.0||1.4|
|EPS 5-yr. historical growth rate (%)||17.4||-0.6|
|EPS 3-5 yr. estimated growth rate (%)||10.8||12.0|
|Market cap. ($ million)||226.0||428.0|
|Relative strength vs. S&P (S&P=0) (%)||-9.0||7.0|
|Data as of March 26, 2010.|
The minimum dividend yield required by the Rea-Graham screen is currently 2% (66.7% of the corporate bond rate of 3%). The median dividend yield among those companies passing the screen is currently 3.2%, while the typical exchange-listed stock does not pay a dividend.
The current Rea-Graham stocks have seen solid earnings growth over the last five years, with a median annual growth rate of 17.4%. By comparison, the median growth rate for exchange-listed stocks was a decline of 0.6% per year.
Although we omitted over-the-counterstocks from consideration, the companies passing the Rea-Graham screen are still relatively small, with a median market capitalization of $226 million. Meanwhile, the typical exchange-listed stock has a market cap of $428 million.
Lastly, the stocks passing the Rea-Graham screen have struggled a bit over the last year, underperforming the S&P 500 by 9% over the last 52 weeks. In contrast, exchange-listed stocks have outperformed the S&P 500 by 7% over the last 52 weeks.
Stocks on the List
Table 2 lists the 22 companies passing the Rea-Graham screen as of March 26, 2010. These companies are ranked in descending order by earnings yield.
|Company (Exchange: Ticker)||
|EarthLink, Inc. (M: ELNK)||896.9||31.3||6.7||0.0||2.0||30.2||82.3||Internet serv|
|USA Mobility (M: USMO)||288.0||22.7||7.8||0.0||3.5||38.4||90.8||communications|
|Daxor Corporation (A: DXR)||47.2||15.2||5.4||0.8||1.0||82.6||58.4||medical devices|
|Air T, Inc. (M: AIRT)||28.3||13.9||2.8||0.0||1.4||17.7||93.3||air cargo & support|
|Kewaunee Scientific (M: KEQU)||35.9||11.1||2.9||0.3||2.7||23.0||74.2||lab furniture|
|Ampco-Pittsburgh (N: AP)||262.3||10.6||2.8||0.0||nmf||64.5||47.7||steel & pumps|
|Span-America Med Sys (M: SPAN)||49.5||9.7||2.2||0.0||4.5||17.4||99.9||foam prods|
|Hawkins, Inc. (M: HWKN)||241.0||9.6||2.4||0.0||4.5||32.5||88.5||specialty chemicals|
|Ambassadors Group (M: EPAX)||210.9||9.3||2.2||0.0||4.2||6.3||54.3||educational travel|
|Kaiser Aluminum Corp. (M: KALU)||774.2||9.2||2.5||0.8||6.4||18.4||49.1||aluminum prods|
|CCA Industries, Inc. (A: CAW)||47.0||9.0||5.1||0.0||1.5||-8.9||56.2||beauty prods|
|Espey Mfg & Electron (A: ESP)||38.4||9.0||4.4||0.0||1.8||22.6||85.5||elec power supplies|
|Ecology and Environ (M: EEI)||60.7||8.6||2.9||0.5||1.9||15.8||86.5||environ’l consult|
|Hillenbrand, Inc. (N: HI)||1,370.7||7.8||3.4||0.0||nmf||na||88.8||funeral servs|
|Cherokee Inc. (M: CHKE)||159.2||7.4||8.4||0.0||15.2||-1.0||48.9||licensed apparel|
|Buckle, Inc., The (N: BKE)||1,711.0||7.4||2.2||0.0||13.0||25.4||82.8||casual apparel|
|Wayside Technology (M: WSTG)||44.5||7.0||6.5||0.0||2.5||-17.0||85.1||IT software|
|Terra Nitrogen Co. (N: TNH)||1,461.2||6.9||8.3||0.0||nmf||17.3||45.6||fertilizer prods|
|Cato Corp. (N: CATO)||631.0||6.8||3.1||0.0||3.6||5.4||92.1||women’s apparel|
|Collectors Universe (M: CLCT)||80.7||6.7||9.5||0.0||7.6||-20.5||96.0||collectible grading|
|Weis Markets, Inc. (N: WMK)||968.6||6.5||3.2||0.0||7.0||2.0||89.4||retail food stores|
|Automatic Data Proc’g (M: ADP)||22,333.8||6.1||3.1||0.7||20.7||11.7||96.2||business servs|
Exchange Key: A=American Stock Exchange; M=NASDAQ; N=New York Stock Exchange.
Source: AAII’s Stock Investor Pro/Thomson Reuters. Data as of March 26, 2010.
EarthLink, Inc.), an Internet service provider primarily offering dial-up Internet access to consumers, has the highest earnings yield among the passing companies at 31.3%. Technology trends, for the most part, have passed EarthLink by as more consumers shift to high-speed Internet. The company faces well-entrenched competition with much deeper pockets as it attempts to shift its focus to delivering broadband access. As a result, the company has a current price-earnings ratio of 3.2 (1 ÷ 0.313 = 3.195). Interestingly, the company has been able to growth its earnings at an average annual rate of 30.2% a year over the last five years as the company has focused on deep cost-cutting measures.
Collector’s Universehas the highest dividend yield among the passing companies at 9.5%, based on an indicated annual dividend of $1.00 per share. The company provides authentication and grading services to dealers and collectors of high-value coins, trading cards, event tickets, autographs, memorabilia and stamps. However, the company’s dividend policy has been erratic since it began paying a dividend in 2006: After paying out $0.91 a share in dividends in 2008, it only paid $0.23 per share in dividends in 2009.
What It Takes: The Rea-Graham Screen
- The earnings yield is greater than or equal to twice the average AAA corporate bond rate (alternatively, the price-earnings ratio is less than or equal to one-half of [100 ÷ the average AAA corporate bond rate])
- The dividend yield is greater than or equal to two-thirds the average AAA corporate bond rate
- The debt-to-equity ratio is less than or equal to one
- Excludes stocks in the financial sector
- Excludes REITs
- Excludes stocks that trade over the counter )
- Excludes ADRs and other foreign companies
Benjamin Graham and James Rea were firm believers in finding solid companies trading at low valuations. However, they also believed that any stock screening system needed to be easily understood and implemented. Likewise, in order to protect themselves from individual company risk, they diversified their portfolio across 30 or more companies. For many investors, such a simplistic approach to stock-picking and investing in such a large number of stocks is unappealing. However, the wealth of academic research that seemingly advocates such value-oriented investing methods and the success of Graham himself over his career—as well as followers of his such as Warren Buffett—make it difficult to dismiss this approach as outdated.