Tuesday, October 8, 1996
Cheating the
Dow
by MF Cormend
Investors!
I'd like to announce officially in this forum a powerful new strategy that can lead to annual returns of 50% or more. Furthermore, this strategy takes less than one hour a year of your time and is extremely safe because it involves investing in only the largest blue chip companies in America, those included in the Dow Jones Industrial Average!
The strategy is up 34% so far this year! It easily beats the S&P 500 (up 12%), the Dow Jones Industrial Average (up 15%), and far and away surpasses even any Beating the Dow variation yet devised. Imagine doubling Beating the Dow!
The success of the system is based on a fact well known among a handful of the most successful Wall Street traders, those dealers who "make the markets." Basically, it is known that those stocks that end the day with quotes listed in eighths (e.g., 76 1/8 or 43 5/8) do poorly in the medium term, as measured in months. The theory behind this observation is complex, but for the sake of argument, let's take these experienced traders at their word.
How do we take advantage of this statistical quirk? Simple! Let's use the 30 Dow stocks as of January 1 and throw out all the stocks whose quoted price ends in an eighth. These are the under-performers. Then, borrowing a page from Michael O'Higgins, we list these stocks from cheapest in price to the most expensive. The five lowest priced of these "non-eighths" stocks make up our new stock portfolio for the year. I'll call the system "Cheating the Dow" (CTD), and the five CTD stocks for 1996 are listed below, along with their year-to-date gains (as of 9/20/96):
Woolworth +66.4%
Union Carbide +21.2%
Sears +20.0%
Allied Signal +33.5%
Caterpillar +29.7%
Total return for 1996? How's 34.2%!
If any investor would like more info about this exciting strategy, pl... STOPSTOPSTOPSTOPSTOPSTOPSTOPSTOPSTOPSTOPSTOPSTOPSTOPSTOP
Sound tempting? Please don't be fooled!
The Cheating the Dow strategy is just a figment of my imagination, although the data incorporated are very real. I developed it to illustrate some of the pitfalls commonly seen in investment advertising, as well as what Fools should look for when deciding to put their money in any investment strategy.
The system was created simply by observing what stocks have done well this year, in retrospect. I then searched for any combination of identifying features for these stocks that could distinguish those that had done particularly well. Yields, prices, trading symbols, vowels in the names of the company, whatever it took to segregate five stocks which as a group did extraordinarily well. The "eighths rule" combined with the price gave just the result I wanted.
Although these data were calculated using only 1996 data, there is no reason in the world that the same process could not have been used retrospectively to give incredible appearing results over longer time periods (e.g., 20 or 30 years), thus appearing to give even added legitimacy to a particular system. Statisticians have long recognized that when trying to find meaningful results in data, the validity of the findings is inversely related to the number of variables that are analyzed. These multiple comparisons, or "data dredging," are the heart of many Wall Street aphorisms such as the Super Bowl Theory, in which it is said that the winning team determines the direction of the market for the year. In fact, based on similar false statistical analyses, it was "scientifically proven" in a medical journal that a certain type of treatment for heart attacks was only beneficial if one was born under the sign of Leo.
What are the lessons of Cheating the Dow?
1) Any investment strategy should have some basis in sound financial principles.
2) The methodology used to develop the strategy should be clearly stated.
3) The longer the back-tested results the better, everything else being equal.
4) A successful system should also work when used prospectively.
For those of you interested in the current Cheating the Dow stocks, please DON'T send me $79.95 and in return I WON'T send you the first 2 months of the Cheating the Dow Newsletter ABSOLUTELY FREE !
Transmitted: 10/8/96