The Foolish Four has been under fire for not living up to our expectations, and we Fools have been under fire for not realizing sooner that there were possible problems with the way the strategy was developed -- and for promoting it as a simple, almost foolproof way to invest. Hindsight is 20-20, but we aren't yet to the point where we can apply hindsight accurately. We are working to modify our advice, learn from our mistakes, and do more rigorous testing of the Foolish Four formula.
By
When the Motley Fool Investment Guide was published in 1996, Tom and David spent a lot of pages enthusiastically recommending Michael O'Higgins' Beating the Dow strategy, and they introduced our own version, the Foolish Four, which looked even better. A lot of what was written then has come back to haunt us recently -- primarily the idea that the Foolish Four is the logical next step after index fund investing, and that an investor need go no further to reap market-crushing, long-term rewards. Life, and investing, is just not that simple.
The Investment Guide is being revised. The new edition, which won't be hitting the bookstores for several months, expresses our disappointment in the way the Foolish Four has played out. The strategy is put in context as part of a valuable learning process that led eventually to the birth of the Workshop and to the mechanical investing concept, as developed by the Workshop community.
I am hopeful that those critical of the Fool's former championing of the Dow strategies as one-stop investing will also put the book in context. When the Investment Guide was written, Beating the Dow had been out for five years, and for four of those five years the strategy had trounced the Dow. (Ironically, BTD lost to the Dow badly in the year it was first published. Remember the Gulf War recession?) The post-publication performance, combined with the 17 years of outstanding historical returns in O'Higgins' first book, made a pretty convincing case. To us, anyway. Here's how things looked then:
Dow 30 BTD5 Fool 4
1990 -9.14% -17.34% -17.61%
1991 30.36% 59.14% 34.81%
1992 11.04% 20.19% 30.24%
1993 17.91% 30.49% 30.26%
1994 3.70% 6.10% 7.38%
CAGR* 9.97% 16.96% 15.14%
*Five-year compound annual growth rate
Note: All returns are based on portfolios started the first trading day of each year.
Five years of what statisticians call out-of-sample returns may not have constituted ironclad proof, but they went a long way toward dispelling worries that this was one of those strategies that, once publicized, promptly fails to work. For another take on this, see the recent New York Times article "In the Data Mine, There Is Seldom a Pot of Gold." (Who writes those headlines?) You will have to register to follow the link, but it's free.
So, in 1995, high-yield/low-price Dow strategies looked pretty darn good. Here's how things looked for the next five years:
Dow 30 BTD5 Fool 4
1995 36.69% 30.13% 47.05%
1996 24.32% 26.62% 26.56%
1997 22.33% 17.70% 19.49%
1998 15.95% 12.49% 15.64%
1999 20.57% -7.60% 21.47%
CAGR 23.79% 15.05% 25.58%*
*Note: Although the current version of the Foolish Four strategy beats the Dow for those five years, it doesn't beat the S&P 500, which averaged a remarkable 28.57% annual return from 1995 through 1999.
Don't get distracted by the high overall returns up there. What counts is how the strategies did relative to the Dow. That's not so impressive. Although Foolish Four portfolios started in January have kept pace with the Dow, they have hardly been turning in the kind of Dow-doubling performance that was expected. And the underperformance of Beating the Dow causes us to wonder if the whole high-yield and low-price thing is really such a great idea.
It is that long-term performance that has us worried, by the way, not this year's dismal state. Some other things that have us concerned are the variability in the returns of portfolios started on different days, the changing nature of the Dow, the lack of apparent interest in dividends, the market's devotion to growth stocks, the need to demonstrate statistical validity, and global warming. (Still awake?)
All of this must be making some Foolish Four investors feel very insecure. I don't blame you. It's a rare day that goes by that someone doesn't ask me about the future of the Foolish Four. So, for the record, let me state that I think it is a good strategy that is going through a bad period, very much like it has done at times in the past.
That's one Fool's opinion.
On the other hand, we have no proof that the strategy works in any kind of absolute statistical way. (I thought I had statistical evidence, although not proof, but that was back when I thought statistics was actually fairly simple, and even kind of fun.) We have a history of outperformance that could be the result of chance -- or could be an association that was valid in the past, but no longer holds. We also have this year's abysmal performance, and last year's abysmal performance of portfolios that included Sears <% if gsSubBrand = "aolsnapshot" then Response.Write("(NYSE: S)") else Response.Write("(NYSE: S)") end if %>, Goodyear Tire & Rubber <% if gsSubBrand = "aolsnapshot" then Response.Write("(NYSE: GT)") else Response.Write("(NYSE: GT)") end if %>, or DuPont <% if gsSubBrand = "aolsnapshot" then Response.Write("(NYSE: DD)") else Response.Write("(NYSE: DD)") end if %>. It hasn't been pretty.
All of this does not make a firm foundation for a strategy that makes up a large part of anyone's investment portfolio. Today, with the benefit of hindsight, I think the Foolish Four is far more appropriate as a value component of a much more diversified portfolio. Does saying that mean I "really" think it doesn't work anymore, or that it will never work again, or that I agree it was a product of data mining and was never valid to begin with? Nope. I also know that any or all of those things may turn out to be true.
The fact is that it takes time to figure out the answers to those questions. While the second chart up there shows that the Foolish Four strategy has barely beaten the market for the past five years (and if you added in this year, it would be losing to the Dow and the S&P 500 over the entire time period since 1995), that five-year period looks a lot like some other five-year periods in the past.
It may turn out that the critics who say that the strategy was simply a random association of variables in a vast database are correct. The only way I know to settle the question is to get a different database and test the strategy on a different set of stocks over different time periods, and we are doing that. It's taking far longer than I had hoped, due to problems with the database and that darn 24-hour-a-day limit, but it is being done and will be released just as soon as we are reasonably sure we have done things right.
In the meantime, what do you do?
The same thing you've always done. Weigh the evidence, and decide what is right for you.
If you are interested in seeing more details about what the very vocal opposition believes, check out our discussion board. (Wear your fire-retardant gloves when clicking in there.) They make the case for the "data mining" school of thought very well. For my rebuttal, see the archives for May 1999. The discussion starts on May 10 and goes on and on, finally winding up around May 21, which might actually be the best place to start if you aren't masochistic enough to wade through the whole thing.
What does all this mean for the Workshop? We are running this column jointly in the Workshop and Foolish Four areas because the Foolish Four is the original mechanical investing strategy. The way it was developed is a far cry from the work going on in the community now, but even so, I see the Foolish Four as a cautionary tale.
Although some of our Workshop strategies have been far more thoroughly tested than Beating the Dow or the Foolish Four were when they were first embraced by Fools, we need to remember that any mechanical strategy can pull a Foolish Four, whether they are "valid" or not. Markets change. Mechanical strategies can underperform at random.
If you look closely at the backtests of even our strongest mechanical strategies, there are almost always periods during which the strategy lost money and/or lost to the market. One of the most popular strategies, the monthly version of Relative Strength-26 weeks, lost to the market in 1969, 1972, 1975 (one of the greatest rally years ever, and it lost 8% compared to the S&P's gain of 37%!), 1984, 1992, and 1997. Plus, the strategy tied the market two years running -- in 1988 and 1989. (Data courtesy of Jamie Gritton's Backtest Engine. Thanks, Jamie.)
Anyone investing in just one Workshop strategy or using several strategies that use similar criteria needs to think seriously about diversifying into complementary strategies -- yes, maybe even the Foolish Four as a value strategy to balance the high-growth strategies that are so popular right now. And Foolish Four investors who are still attracted to the idea of mechanical investing should put in some time educating themselves on other strategies.
Yeah, I know. We promised you that investing this way would only take 15 minutes a year. We were wrong about that, too.
Fool on and prosper!
Related Links:
The Value of Value Stocks, Foolish Four Portfolio, 8/7/00
Why I Believe in the Foolish Four, Foolish Four Portfolio, 8/4/00
The Fuss Over the Foolish Four, Foolish Four Portfolio, 7/31/00
Foolish Four Strategy Still Valid?, Foolish Four Portfolio, 7/28/00