|
PureBytes Links
Trading Reference Links
|
Great !!
I saved this message to read it carefully.
--- In amibroker@xxxxxxxxxxxxxxx, Yuki Taga <yukitaga@xxxx> wrote:
> Hi DIMITRIS,
>
> Friday, February 27, 2004, 4:52:04 PM, you wrote:
>
> DT> Yuki, the question is : Could we collect all these interesting
> DT> comments, code them and add their wisdom into the new 2004timing
> DT> ? We have 3 more months untill the end of May2004 when this code
> DT> will be activated. There are some factors closely related to the
> DT> timing of this market, seasonal or not. How many are known in
> DT> advance ? We may [and we should] code them. Just think about
if
> DT> we could prepare a list of events/factors.
>
> That is the rub, isn't it? I can say with fair certainly on the
> stocks that I trade that there is a beginning of the month bias
> toward the main trend. This is a measurable bias, on calendar days
1
> through 4, and it is even stronger given that our market here never
> opens on January 1, 2, or 3, nor on November 3. So the bias is even
> stronger given that those days are missing from the sample data. (I
> am not sure exactly how to code the first 4 *trading* days of a
> month, rather than the first 4 calendar days.)
Whenever you donīt know or you are not sure, the solution may be
SIMPLE : Just ask and here it is :
fourtradingbars=Sum(Month()!=Ref(Month(),-1),4)==1;
Plot(C,"",1,64);
PlotShapes(shapeCircle*fourtradingbars,colorRed);
A red circle will be plotted the first 4 trading bars of the month.
The [binary] condition "fourtradingbars" will be true for these days
and false for the rest.
Enjoy the power of AFL !!
Dimitris
>
> As you may have gathered, while I was stunned at the initial success
> of this timing model in the second half of last year, I am not sure
I
> could ever actually commit money on such a model. It is *very*
> interesting, but somehow predetermined buy and sell points seem
> destined to fail after some period of time. Probably, some external
> events (like unwinding cross share holdings, or a yen correction)
> cause them to fail eventually -- they just interrupt the cycle.
> Perhaps what is needed is a more frequent update to the timing
model.
> You suggested that after one year, it probably gets stale. Maybe it
> gets stale more quickly.
>
> How does a timing model test on in-sample data, where the timing is
> adjusted more frequently? For example, you ran a baseline from 1/1
> through May last year to get the current model. What if, as the
> current model is running, another baseline is being established from
> June through November, for use this year? Maybe it doesn't change
> anything; maybe it does.
>
> Another suggestion: How does the system do if we do not initiate a
> trade until we have both a H and L in the direction we are going?
For
> example, Thursday morning was the short signal. But, instead of
> taking it on the open we wait to see if Thursday's H takes out
> Wednesday's H. It did, so we don't enter the short until we have a
> lower H and a lower L. I have used some other intermediate trend
> indicators in this manner, and it seems to be a decent fail safe.
> This short may end up working (really, it looks bad, but then I've
> seen it all come to pass, including that which I thought
impossible),
> but taking it on Thursday morning right now would have you popping
> lots of pain medication by now. One would have to have *extreme*
> confidence to stay with the trade here, so a delayed entry might be
> something to consider. For a good call, the entry would be delayed
> only by one day, which should not matter much. In a bad call, it
> could save a lot of cash and grief.
>
> Yuki
Send BUG REPORTS to bugs@xxxxxxxxxxxxx
Send SUGGESTIONS to suggest@xxxxxxxxxxxxx
-----------------------------------------
Post AmiQuote-related messages ONLY to: amiquote@xxxxxxxxxxxxxxx
(Web page: http://groups.yahoo.com/group/amiquote/messages/)
--------------------------------------------
Check group FAQ at: http://groups.yahoo.com/group/amibroker/files/groupfaq.html
Yahoo! Groups Links
<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/amibroker/
<*> To unsubscribe from this group, send an email to:
amibroker-unsubscribe@xxxxxxxxxxxxxxx
<*> Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/
|