Looking at the Evidence
Do portfolios of stocks with low priceearnings ratios outperform the market? The answer to this question is central to this chapter, and you will look at the performance of stocks with low PE ratios over the last few decades in this section.
Ben Graham and Value Screening
Many value investors claim to trace their antecedents to Ben Graham and to use the book Security Analysis that he coauthored with David Dodd, in 1934 as their investment bible.^{[4]} It was in the first edition of this book that Ben Graham put his mind to converting his views on markets to specific screens that could be used to find undervalued stocks. While the numbers in the screens did change slightly from edition to edition, they preserved their original form and are summarized below:

Earningstoprice ratio that is double the AAA bond yield

PE of the stock less than 40% of the average PE for all stocks over the last five years

Dividend yield greater than twothirds of the AAA corporate bond yield

Price less than twothirds of tangible book value^{[5]}

Price less than twothirds of net current asset value (NCAV), where net current asset value is defined as liquid current assets including cash minus current liabilities

Debtequity ratio (book value) less than 1

Current assets greater than twice current liabilities

Debt less than twice net current assets

Historical growth in EPS (over last 10 years) greater than 7%

No more than two years of declining earnings over the previous ten years
Note that the first screen is a priceearnings ratio screen. Only stocks with low priceearnings ratios would have a chance of passing this screen. It is interesting that many of the remaining screens are designed to eliminate those stocks that have low PE ratios for the wrong reasons: low growth and high risk.
How well do Ben Graham's screens work when it comes to picking stocks? Henry Oppenheimer studied the portfolios obtained from these screens from 1974 to 1981 and concluded that you could have made an annual return well in excess of the market.^{[6]} Academics have tested individual screens—low PE ratios and high dividend yields to name two—in recent years and have found that they deliver higher returns. Mark Hulbert, who evaluates the performance of investment newsletters, found newsletters that professed to follow Graham did much better than other newsletters. The only jarring note is that an attempt to convert the screens into a mutual fund that would deliver high returns did fail. In the 1970s, an investor named James Rea was convinced enough of the value of these screens that he founded a fund called the ReaGraham fund, which would invest in stocks on the basis of the Graham screens. While it had some initial successes, the fund floundered during the 1980s and early 1990s and was ranked in the bottom quartile for performance.
Low PE Stocks versus the Rest of the Market
Studies that have looked at the relationship between PE ratios and excess returns have consistently found that stocks with low PE ratios earn significantly higher returns than stocks with high PE ratios over long time horizons. Since some of the research is more than two decades old and the results vary widely depending upon the sampling period, it might be best to review the raw data and look at the longest period for which data is available.
In Figure 3.2, you begin by looking at the annual returns that would have been earned by U.S. stocks categorized into ten classes according to PE ratios from 1952 to 2001. The stocks were categorized by PE ratios at the start of each year, and the total return, inclusive of dividends and price appreciation, was computed for each of the ten portfolios over the year.
Figure 3.2. PE Ratios and Stock Returns: 1952–2001
On average, the stocks in the lowest PE ratio classes earned almost twice the returns of the stocks in the highest PE ratio classes. To examine how sensitive these conclusions were to how the portfolios were constructed, you can look at two constructs. In the first, equally weighted portfolios were created, and an equal amount of money was put into each firm in each portfolio. In the second, more was invested in the firms with higher market value and less in the smaller firms to create valueweighted portfolios. The results were slightly more favorable with the equally weighted portfolio, with the lowest PE ratio stocks earning an average annual return of 24.11% and the highest PE ratio stocks earning 13.03%. With the valueweighted portfolios, the corresponding numbers were 20.85% and 11%, respectively. In both cases, though, low PE stocks clearly outperformed high PE stocks as investments.
To examine whether there are differences in subperiods, let's look at the annual returns from 1952 to 1971, 1972 to 1990, and 1991 to 2001 for stocks in each PE ratio portfolio in Figure 3.3. Again, the portfolios were created on the basis of PE ratios at the beginning of each year, and returns were measured over the course of the year.
Figure 3.3. Returns on PE Ratio Classes: 1952–2001
Firms in the lowest PE ratio class earned 10% more each year than the stocks in the high PE class between 1952 and 1971, about 9% more each year between 1971 and 1990, and about 12% more each year between 1991 and 2001. In other words, there is no visible decline in the returns earned by low PE stocks in recent years.
Thus, the evidence is overwhelming that low PE stocks earn higher returns than high PE stocks over long periods. Those studies that adjust for differences in risk across stocks confirm that low PE stocks continue to earn higher returns after adjusting for risk. Since the portfolios examined in the last section were constructed only with stocks listed in the United States, it is also worth noting that the excess returns earned by low PE ratio stocks also show up in other international markets.