The Keystone and Keystone EPS have been discussed in previous workshop articles. Today we look at Key 100 and compare the three versions.
The Keystone 100 screen is very similar to the original Keystone. It selected the largest domestic market cap stocks from among the top 400 most timely stocks as rated by Value Line. The top 30 stocks by market cap were selected and those stocks were then ranked by total return over the past 6 months. Why 30 stocks? Well, creator Robert Sheard was looking for a high growth alternative to the Dow -- which consists of 30 stocks.
The Keystone EPS starts with the original Keystone 10 stocks and ranks them by "earnings per share growth" over the past year. Many investors feel that including a fundamental stock selection criteria like earnings growth makes the Keystone screen more reliable.
But Foolish Workshop contributor Powerphil wondered how the Keystone would work if it wasn't limited to just 30 high capitalization stocks from the U.S. He tested a version that included foreign companies, as long as they were traded on the New York Stock Exchange, American Stocks Exchange or Nasdaq. He changed the cutoff point from 30 to the 100 largest companies.
The steps for doing the Keystone 100 using Value Line for Windows Software are as follows:
1. Select only stocks with a Timeliness Value of 1 or 2 (this will give you a list of 400 stocks)
2. Eliminate all stocks not traded on the NYSE, AmEx or Nasdaq exchanges
3. Sort by market cap, highest to lowest, and select the top 100 stocks
4. Sort by "Total Return - 26 Weeks," highest to lowest
5. Buy and hold for one year either the top 5 or top 10 stocks
Three versions of Keystone can make it tough to decide which one to use so let's compare returns. The following returns are the compounded average growth rates for five stock versions of each portfolio tested from 1987 through 1998. Twelve portfolios were tested for each strategy, each portfolio starting/renewing in a different month.
Month Keystone KeyEPS Key 100
Jan. 27.5 31.6 33.2
Feb. 30.2 29.5 40.1
Mar. 18.3 19.8 29.4
Apr. 23.9 20.4 24.5
May 20.3 19.8 21.2
June 21.2 23.8 22.3
July 20.6 21.5 19.6
Aug 26.3 28.4 14.7
Sep. 21.1 20.8 28.6
Oct. 17.5 18.7 26.1
Nov. 25.4 23.3 25.4
Dec. 33.5 34.1 29.7
Avg. 23.8 24.3 26.2
The data shows that overall and on a month by month basis, the Keystone-EPS beats the original strategy, but not by very much. In contrast, the Keystone 100, which uses a larger pool of stocks to select from, beats them both. The Keystone 100 does best in seven starting months, compared to four for the Keystone EPS with a tie in November between the original Keystone and Keystone 100.
The numbers also show a seasonal effect with December, January, and February outperforming most other months, although one should probably take the differences between those three with a large dose of salt. With a 12 year backtest, the data is somewhat vulnerable to the effects of chance when it comes to something like the difference between a January start date and a February start.
The differences between the three are not major, although it is rather difficult to justify the original Keystone when compared to the results of the Key100. The Key100 seems to be a significant improvement on an old favorite.
Until next time, Fool On!