In search of information content: portfolio performance of The 100 Best Stocks to Own in America

Financial Services Review, Summer 2005 by Anderson, Randy I, Loviscek, Anthony L

4. Results

We are able to obtain the necessary data for all 100 stocks in each edition, a total of 600. To apply Eq. (2) of the EGP approach, we need to estimate an expected return for each of the 600 stocks. To do so, we follow Ritter (2002) and use the monthly return that annualizes to the 10-year return that Walden provides for each stock. For example, if Walden reports an annual return of 24% for a stock, we use 1.8088%. We also note that the 10-year historical return, a conspicuous number that likely influences the individual investor's choice of stocks, is one of Walden's major screening criteria and serves as a reasonable and objective number on which to base an expected monthly return.

To estimate Eq. (1), the single-index model, we use the Beta Book of Ibbotson Associates (2002) as a guide. We use five years of monthly historical returns for each stock and the 30-day U.S. Treasury bill yield as the risk-free rate of return. ' We include betas significant at the 10% level. Across the 600 regressions, we find that only five betas do not meet this criterion, and more than 95% are significant at the 5% level.

Four acquisitions occurred during the testing periods. To handle them, we compute the monthly returns for the stock in question up to the time in which it was no longer traded, and assume that the portion of the portfolio initially allocated to that stock now holds cash that earns the risk-free rate of return. For example, Pfizer completed its acquisition of WarnerLambert, which is one of Walden's selections, in June of 2000. We compute the rate of return of the stock up to the final day of trading, and then assume that the risk-free return is earned through October 1, 2001, the remainder of the testing period.

As some insight into the performance of Walden's selections, Table 1 provides the monthly excess returns (i.e., in excess of the 30-day Treasury bill return) and the risks, as measured by standard deviations, for all 100 stocks and the S&P 500 for the tracking period of each edition. The range of the excess returns on the 100 stocks is from -0.242% (sixth edition) to 1.726% (fourth edition), smaller than that of the S&P 500, which is from -1.262% (sixth edition) to 1.984% (fourth edition). The excess returns on the 100 stocks from the first, fifth, and sixth editions are greater than those of the S&P 500. Overall, the average monthly return for the 100 stocks, based on a geometric mean, is 0.721%, which exceeds that of the S&P 500 at 0.619%. At this level, the results suggest that Walden's selections have some potential to outperform the S&P 500.

The risks of the 100 stocks range from 2.472% (third edition) to 5.195% (first edition), a little bit wider than the range of risks of the S&P 500, 2.604% (third edition) to 5.209% (fifth edition). However, the risks of the 100 stocks are lower than those of the S&P 500 for the third through the sixth editions. Thus, the 100 stocks from the fifth and sixth editions have both higher returns and lower risks than those of the S&P 500. Overall, in terms of the sizes of the returns and risks, the 100 stocks appear to offer a small advantage over the S&P 500. Before drawing firm conclusions, however, we turn to more detailed results, as shown in Tables 2-4.