|
PureBytes Links
Trading Reference Links
|
Clement,
In May of 1997, you published to RealTraders a customized algorithm of
Gann's Swing. I transferred it into my TradeStation and am currently
trying to apply it. I am wondering if you have done any more research
with this study, and if you have would you like to give us an update.
The market that I am looking at is the May Soybean; the market made a
new swing low on Friday, March Sixth. The low of Friday should have
reestablished the down trend, for SK now has a series of lower swing
highs and lower swing lows. I have attached a gif or the May Soybean;
your Gann Swing study is the thin black line. Would you consider a
short trade if Friday's swing low is taken out sometime next week?
Dennis
schong wrote:
>
> Rters,
>
> Here are 2 fallacies which I would like to highlight
> regarding change in trend(CIT) techniques based on
> my experience :-
>
> i) "Trade with the trend". Take only the CIT signals that is
> with the trend
>
> Inferior CIT techniques hides under the cover of the
> statement above. The problem with it is because
> most CIT can be +/- 1 day out and if you were to
> apply the trend as the filter you are actually trading
> based on reversals in the direction of the trend.
>
> The CIT technique if studied closely would not have
> changed the decision to enter a trade. Hence, taking
> trades based on taking out the high or low of the
> previous bar is misleading trading setup.
>
> It should be note that most CIT reversals technique
> signals a trade when the market takes out the high or
> low of the previous 1 period bar extreme or variant of
> it in the opposite direction.
>
> ii) A random date generator provides CIT forecast with
> result matching or bettering the CIT technique
> publicly discussed, .eg randomly generating 40 dates
> per year.
>
> Assuming that the trader is totally discipline, .eg
> able to execute his trade flawlessly then this is my
> explanation of the paradox.
>
> 1 ) The total number of trades is generally reduced
> by incorporating CIT techniques whether or not the
> CIT has any predictive value.
>
> 2 ) Most of the money is lost when the market is
> in a congestive phase, .eg making 1 or 2 bar
> reversals in succesion and/or small blips(runs).
>
> By using a truely random date generator the number
> of times the trader is faced with condition 2) is
> greatly reduced as a percentage to the number of
> total trades as compared to before when the trader
> would take trades based on every 1 or 2 bar
> reversals and hence more profitable.
>
> This is based on my observation that markets do not
> go into several congestive phase in successions and
> with the randomly generated dates spaced to be at
> least as large as the normal period of the congestive
> phase would tend to increase the bottom line within
> the context of the trader's time frame.
>
> For a CIT technique to truely work the following have to be
> observed :-
>
> i) You must be able to indicate whether a CIT is going
> to be a top or bottom consistently in advance OR
>
> ii) You have a different way of discerning the direction of
> the CIT OR
>
> iii) You must be able to determine the magnitude of the CIT
> move consistently, eg. you can say with high degree of
> confidence whether a CIT is going to MAJOR or at least
> going to x points OR
>
> iv) You can observe certain consistent market characteristics
> around the CIT period which in a non CIT period would
> be consistently invalid OR
>
> v) You can consistently say the CIT date is going to
> come in exact on the date forecasted. No allowance
> for +/- 1 day.
>
> Apply any of the five test to all CIT techniques whether
> they are based on mathematics, geometry, natural law,
> numerology or astrology and you will know whether the CIT
> technique have any predictive value.
>
> It is not my intention to provide proof of actual
> techniques meeting the above test(s) but merely to provide
> food for thought for anyone seriously wanting to
> incorporate CIT technique to their advantage.
>
> I'll close this off by saying "An inferior CIT technique
> will consistently get you into the wrong side of the
> trade".
>
> Clement
--
MIG's MARKETS: http://homepage.macomb.com/~mig/mig.htm
Attachment Converted: "c:\eudora\attach\SK.gif"
|