|
PureBytes Links
Trading Reference Links
|
Someone wrote:
> Is anyone else getting torqued off at the stupid articles - er, I mean
> infomercials - from Tom DeMark and T.J. DeMark that have appeared in
> the past several issues of Futures? I suspect that the DeMarks are
> competent technical analysts - my point is NOT that their methods are
> flawed. My point is that from the articles, there is no way of knowing
> WHAT their methods are! Everything refers you to their stupid,
> proprietary indicators, all of which, embarassingly enough, are named
> "TD - something" (some psychiatrist could have a field day analyizing
Two months ago I had a very frustrating experience trying to program a
Demark Indicator. Included in TS4.0 is the Thomas Demark expert
indicator, which is locked. If you look at the help index of TS it
explains that this "expert" indicator really displays the REI (Range
Expansion Index).
I happen to have some past reprints of Demark's articles in Futures, so I
programmed in the REI, displayed in on a chart, and ***** -- It didn't
quite match what the TS "expert" indicator displayed. So I played around
some with the inputs - the lookback period and "bars ago" comparison.
Still no match.
So I went to his book where he spends several pages on the indicator, only
to discover that the book calculation of REI was different from the
Futures article. So I changed the coding according to what I interpreted
the book to say, and there was still no match. Looking more closely at
the language Demark uses to describe the indicator, I realized that it is
fundamentally ambiguous in some places. Then he says in his book that he
originally used a lookback period for the REI at 8 days, then when that
didn't work anymore, he switched to 5 days.
At this point my opinion is that some people can present a sheer mass of
complexity in the name of wisdom, and this can so influence popular
opinion so as to elevate that person as an exceptional talent. My only
previous experience of DeMark was a superficial reading of his
"Sequential" Indicator, whose basic idea was that if a stock or future
closes down thirteen days in a row, then it is time to buy it, and vice
versa. (This is a simplification of another complex matter). I wondered
what made 13 so special. Was it a mystical number, or was it simply a
result of optimizing past data, or was it simply that if something closes
down between 7 and 19 days in a row that one of those times it is bound to
close up? I talked to a brokerage firm that hypes the Demark indicators,
especially the sequential, and I was told that the sequential no longer
works very well, so now they are using the Combo (TD Combo to be exact).
There are presuppositions in these indicators that seem either purely
arbitrary or a result of optimization, as for example the requirment in
"Sequential Setup" that we compare the current bar to the bar 4 bars
earlier. However, what got me interested in the REI was the filter that
was used to keep the REI indicator from maxing out when a strong trend was
present. I thought this one idea might have merit and be worth
investigation, but the sheer frustration cut that short. I think of Tom
Demark as similar to Edison, a man who tried a thousand ideas before he
found one that worked. But DeMark has not yet reached his thousand.
If anyone else has thoughts on this or would like to collaborate on the
REI, let's hear from you.
If anyone has programmed an REI that duplicates the display of the TS
Demark Expert, please Email it to me so I can stand corrected.
Neal
|