Value investors have done quite well over the past 30 years. It is well known that low price-to-earnings (P/E) stocks, or value stocks, on average, earn higher returns than high P/E stocks, or growth stocks. But this is not the full reason that value investors have done so well.

Sorting stocks by P/E and investing in the low P/E stocks is only part of how value investors choose stocks to invest in. Some low P/E stocks may not be worth investing in as they may have a low P/E because they’re simply bad stocks. How do value investors separate the good stocks from the bad among the low P/E stocks? Value investors value individually each stock to determine its intrinsic value and only invest in the stocks that afford them a satisfactory “margin of safety.”

But this is a very time consuming exercise for the average investor. Is there a strategy – an additional screening to the low P/E stocks – investors can follow which will enable them to identify the low P/E stocks that are worth investing in without having to go through the time consuming exercise of valuing each one?

On average, 39% of all value stocks that I examined had a negative return for the 12 months that followed their allotment into low P/E stocks. Conversely 50% of the growth stocks had on average a negative return for the 12 months that followed their allotment into high P/E stocks.

Here is how I removed the time-consuming step of valuing each one individually by assigning a SCORE to each stock that is based on publicly available financial ratios from historical company information.

By examining only Canadian non-interlisted companies that traded on Canadian Stock Exchanges from May 1, 1985 to April 30, 2009, I divided stocks into value and growth by sorting them into quartiles using trailing P/E ratios.

Quartile 1 was the lowest P/E ratio stocks, or value stocks. Quartile 4 was the highest P/E stocks, or growth stocks. All data are from COMPUSTAT. I zeroed in on non-interlisted Canadian stocks to ensure the results were driven by Canadian investors as opposed to US investors who tend to be marginal traders in Canadian stocks that are interlisted. Moreover, interlisted stocks tend to be larger, more glamorous and well known, stocks that are followed by many analysts and I did not want the results to be driven by the well known size effect.

After I excluded negative P/E ratios and stocks with prices below $1, I ended up with 7,145 annual observations belonging to 1,237 unique companies.

I sorted all firms in the sample by historical market cap, stock liquidity, asset turnover and revenue, EBIT and EPS annual growth rates and, in each case, based on the sorted ranking, I assigned values from 1 to 6 to the sample firms. The smallest market cap and liquidity stocks and highest asset turnover, revenue, EBIT and EPS growth stocks are assigned the value of 1; the largest market cap and liquidity stocks and lowest asset turnover, revenue, EBIT and EPS growth stocks are assigned the value of 6 and so on.

The weighted sum of the above values per stock (where weights are assigned based on regression analysis) forms the basis for a composite (SCORE) indicator. I then ranked all value and growth stocks by the SCORE indicator and formed six value and growth portfolios with SCOREs from low (portfolio 1) to high (portfolio 6). The indicator is constructed to imply that the lower the SCORE the better it is.

I found that there is a negative relationship between the SCORE indicator and annual stock returns for value and growth stocks. For value firms, the lowest SCORE indicator portfolio had a mean annual return of 36.89%, whereas the highest SCORE indicator portfolio had a mean annual return of -11.35%. The corresponding mean annual returns for the growth portfolio were 31.36% and -28.14%, respectively. The median annual returns were consistent with the mean values.

Both value and growth stocks have much better performance when stock selection takes place by focusing on low SCORE portfolios. Whereas the overall sample mean and median annual returns for value firms over 1985-2009 are 16.86% and 8.90%, respectively, value stocks with SCORE values of 1 or 2 have a mean annual return of about 40% and a median annual return of about 28%.

For the growth firms, while the overall sample mean and median returns are 6.32% and 0.00%, respectively, growth stocks with SCORE values of 1 or 2 have a mean annual return of about 30% and a median annual return of about 15%. These are large differences. As a result, an additional screening (based on the SCORE indicator) to the first screening of the value investing process (i.e., only looking at low P/E stocks) adds considerable value to an investment strategy. In fact, a strategy which would involve shorting the high SCORE value stocks and buying the low SCORE value stocks would have beaten the low P/E portfolio by over 30% over the 1985-2009 period. On the other hand, shorting the high SCORE growth stocks and buying the low SCORE growth stocks would have beaten the high P/E portfolio by over 50% over the same period.

*George Athanassakos, **gathanassakos@ivey.uwo.ca**, is a Professor of Finance and holds the Ben Graham Chair in Value investing at the Richard Ivey School of Business, the University of Western Ontario. He is also the Director of the Ben Graham Center for Value Investing (www.bengrahaminvesting.ca).*