Testing BSP
It passes with flying colors!

by Elan Caspi (

[Editor's note: Today we are hearing from one of our newest staff members, Elan Caspi. Elan is no stranger to the readers of our Dow Investing/Foolish Four or Foolish Workshop message boards, though. He has been participating and offering help and guidance to newcomers there for several years. Today, while Ethan Haskel is vacationing in one of the cultural capitals of the world, Elan shares his take on the Beating the S&P strategy.]

EL CERRITO, CA (July 21, 1999) -- Ethan Haskel (Cormend) gave us the BSP Strategy ages ago and has kept us updated on it every week since. I decided to see if I could validate Ethan's results by running a very similar screen using the Value Line Investment Survey as the primary source of data.

You may recall that the original BSP was conducted by starting with the annual "Business Week 50" issue of Business Week magazine, which lists all Standard & Poor's 500 Index companies by industry grouping. The latest annual issue was on March 29, 1999. Ethan used this list to come up with the S&P 30, a kind of Un-Dow list of large market-cap, dividend-paying stocks. For the exact method he used to create the S&P 30, click here. From this list of 30 stocks he took the ten with the highest dividend yield and sorted them by price, lowest to highest. The BSP strategy then selects the stocks ranked 2 through 6. The Business Week issue of March 1998 was the basis for the BSP portfolio begun on January 1, 1999.

BSP has been a successful strategy, beating the S&P 500 almost every year of the last 12 and returning over 20% a year.

My goals in testing a very similar strategy all over again were:

1. To make sure the results are repeatable. What if Ethan made some huge mistake? (He didn't.)

2. To assemble the raw BSP data without having to dig out all of those old issues of Business Week and to have the data in electronic format without copying laboriously from those old issues by hand. For the few people who have old Value Line disks, this is a big advantage.

3. To conduct an "out of sample" test. Many of you will recall the controversy surrounding claims of data-mining against the Foolish Four method. Well, if we come up with very similar results using a different data set, then we have "out of sample" data that may allay some of those concerns abut the BSP and also about the validity of the Dow-based strategies.

I'll call this new variation Beating the Market (BTM) to differentiate it from the original BSP. Here's how BTM was done:

From the first issue of Value Line for each year, select the 30 largest dividend-paying companies. (Note that this isn't coming purely from the S&P 500 Index but from the whole Value Line database of 1700 stocks. Yet, because I selected for size by market cap and so does the S&P 500, I didn't find any stocks in BTM that weren't also in the S&P 500.) Exclude all foreign companies and utilities when creating the list, and allow no more than three companies with the same Value Line Industry Code. These conditions mimic the Dow's stock makeup. No foreign stocks, no utilities, and usually no more than three companies from the same industry.

Once each year's list was complete, I took the 10 stocks with the highest dividend yield, sorted them from lowest price to highest, and selected the stocks ranked 2 through 6. These became the BTM 5 stocks for that year. (This is the traditional Beating the Dow procedure.)

The BTM method is as close to the BSP methodology as one can get using Value Line data. Surprisingly, when I built the stock lists for every year going back 13 years, I found that the BTM picks were quite different from the BSP picks. In fact, out of the top 6 stocks on the list (ranks 1-6), we got an average of 3.75 identical stocks. More than a third of the stocks were different.

The main reason for the difference was timing. BSP is based on a nine-month-old issue of Business Week (the list comes out in March but the stocks for the test were chosen the following January). BTM selects the stocks based on the current market cap. It's not unusual for these large stocks to rise or fall 20% or more during a nine-month period. So when selecting 30 top stocks by market cap, the list changes quite a bit over those nine months, as does the stock's price and, to some extent, its yield.

For the 12 years for which we have BSP data (1987-1998), the average annual return (CAGR) for the strategy was 20.7%. The CAGR of BTM for the same period was 19.9%. The CAGR for the S&P 500 was 17.8%. We have applied the same strategy to a differently selected stock list, yet in both cases the stock strategy beat the market, and the average annual returns for the two lists are within 1 percentage point. This is very strong evidence that the stock selection strategy works and strong evidence against the charge that BSP is a random occurrence resulting from data mining.

        S&P 500     BSP       BTM   
1987     5.23%     13.1%     -1.46%
1988    16.81%     32.1%     54.00%
1989    31.49%     25.8%     20.06%
1990    -3.17%     -4.3%     -3.06%
1991    30.55%     25.5%     36.31%
1992     7.67%      3.9%      6.41%
1993     9.99%     20.2%      7.81%
1994     1.31%      5.6%     11.70%
1995    37.43%     40.0%     44.23%
1996    23.07%     33.5%     28.21%
1997    33.36%     33.0%     25.96%
1998    28.60%     28.9%     22.40%
Mean    18.53%     21.4%     21.05%
CAGR    17.80%     20.7%     19.86%
Fool on.

Note: Ethan's report on the BSP portfolio will return August 4. To see last week's numbers, click here: Superhuman Feats

Today's Stock Lists | 1999 Dow Returns

Call Your Boss a Fool.

 Recent Foolish Four Portfolio Headlines
  12/28/00  Modifying Mechanical Strategies
  12/27/00  Beating the S&P Year 2000 Recap
  12/26/00  After-Hours Quotes
  12/22/00  Why Include the Foolish 4 Port?
  12/21/00  The Value of Community Input
Foolish Four Portfolio Archives »  

07/21/99 Close
Stock  Change   Last
CAT  +   7/16  60.63
JPM  -1  1/8   132.88
MMM  +   9/16  90.00
IP   +   1/4   54.31

                  Day    Month   Year   History
        FOOL-4   +0.27%   1.54%  27.70%  29.60%
        DJIA     +0.06%   0.29%  20.62%  20.14%
        S&P 500  +0.16%   0.48%  12.79%  13.06%
        NASDAQ   +1.08%   2.83%  25.95%  27.68%

    Rec'd   #  Security     In At       Now    Change

 12/24/98   24 Caterpillar   43.08     60.63    40.73%
 12/24/98    9 JP Morgan    105.51    132.88    25.94%
 12/24/98   22 Int'l Paper   43.55     54.31    24.71%
 12/24/98   14 3M            73.57     90.00    22.33%

    Rec'd   #  Security     In At     Value    Change

 12/24/98   24 Caterpillar 1034.00   1455.00   $421.00
 12/24/98    9 JP Morgan    949.62   1195.88   $246.26
 12/24/98   22 Int'l Paper  958.12   1194.88   $236.76
 12/24/98   14 3M          1030.00   1260.00   $230.00

              Dividends Received      $49.99
                             Cash     $28.26
                            TOTAL   $5184.00