Tekboy,
But--and this is what I think Hulbert was getting at--many so called investing "strategies" derived from back-testing are not grounded in theories at all.
I disagree. Hulbert went much, much farther than that. Worse, he did it offering nothing other than speculation and no evidence regarding his acusations. I'll explain, responding to the stuff he wrote.
Hulbert: "I am not suggesting that Beating the Dow [the newsletter] was engaging in data mining simply because it favored high-yielding stocks. But I detect signs of data mining in two of the specific portfolios the letter recommended."
I need to put a spotlight on that. He detects signs. He suspects there might be indicators. He makes no effort to verify or confirm.
Hulbert: "Consider the Five-Stock System, one of the letter's recommended strategies. That system calls for buying the five lowest- priced stocks among the 10 highest-yielding Dow companies. As far as I can tell, there was no basis for this approach other than the fact that the stocks did better in back- testing."
It's irresponsible for Hulbert to write that when there is a 284-page book written by O'Higgins of the same name as the newsletter. In it is a ten-page chapter titled,"Why High Yield and Low Price Work." Yet Hulbert insists there is no basis for the approach "as far as he can tell."
Hulbert very next sentence: "That is a sure sign of data mining, according to Grant McQueen and Steven Thorley, finance professors at Brigham Young University."
What is a sure sign? That Hulbert isn't aware of any basis for the approach?
Hulbert: "Data mining was also evident in another of the newsletter's strategies, the Penultimate Profit Prospect. That approach calls for buying just one stock: the second-lowest- priced issue among the Dow's highest yielders. (The lowest-priced stock did not do as well in back-testing.)"
That strategy wasn't devised by the newsletter unless the newsletter was owned by Michael O'Higgins which I very seriuosly doubt is the case. The phrase, Penultimate Profit Prospect, was coined by O'Higgins in his book from which the name of the newsletter is derived. Moreover, whereas I infer that Hulbert would have us believe that backtesting is the only reason the lowest priced stock might want to be avoided, on page 197 O'Higgins explains exactly why he believes that stock did poorly over the years.
Hulbert: Given these signs of data mining, it is no surprise that these strategies were laggards.
Again, what signs? Did Hulbert document one verifiable example of data mining? No.
Hulbert: "In response to [the Foolish Four's] disappointing performance, Motley Fool is now searching for yet more construction rules."
Please read the stuff at fool.com
If after reading it you believe that modifications were made because of "disappointing performance" or that they were made as a result of data mining, we'll have to simply disagree.
--Mike Buckley |
