|
PureBytes Links
Trading Reference Links
|
Thanks for the explanations and clarifications, Dimitris. Regarding the Turtles, several years ago I plunked down big money to buy the manual from the www.turtletrader.com site. It is a huge document, but there are only 46 pertinent pages in it that describe the methods. It is so poorly written that I found it too difficult to understand. The English is atrocious, and the examples given, of which there are very few, are all different, so it is very confusing to follow the logic. The money management tables are the key to success, not the entries and exits (which are well known, so I really haven't given anything away on the boards here. Tharp presents the basic Turtle system in his book). You don't even need to use their entries and exits if you don't want to as long as you use consistent triggers that recognize the beginning and end of a trend. The Holy Grail part of the system is the complicated money management and pyramiding scheme presented in tables on several pages of the document. Very little explanatory text accompanies those tables, which I found exceedingly difficult to comprehend. So, I laid the book down and forgot about it until I discovered Amibroker and read other books on trading. Now, my interest is piqued again, but I still find the document confusing to read. I've begun coding the system (which I posted last night), but so far it hasn't given satisfactory results in backtesting. There's probably something I'm doing wrong, and I think it has a lot to do with filtering the proper stocks to trade. Al Venosa
>From: "dtsokakis"
>Reply-To: amibroker@xxxxxxxxxxxxxxx
>To: amibroker@xxxxxxxxxxxxxxx
>Subject: [amibroker] Re: QQQ Individual Analysis
>Date: Mon, 24 Jun 2002 06:11:57 -0000
>
>--- In amibroker@xxxx, "Avcinci" wrote:
> > Dimitris,
> >
> > >>When i give EXACTLY individual analysis, you ask "why use it ONLY
>on the QQQ? ". It is strange, isnt it?>>
> >
> > In your very first post on this subject, it appeared, at least to
>me, you were suggesting it for the QQQ. You made no mention of it
>being used for anything else. So, naturally, I had the impression you
>had developed this system for QQQ only.
> >
> > >>[have you seen anything more simple than this ?]>>
> >
> > Yes, the time-proven Turtle breakout system. They go long when the
>highest high of the last 55 days is exceeded (breakout) or go short
>when the lowest low of the last 55 days is exceeded. They exit the
>long position when the lowest low of the last 20 days is exceeded or
>the short position when the highest high for the last 20 days is
>exceeded. That's it. No indicators, no moving averages, no
>stochastics, no oscillators, only the high and low. I don't know of
>any simpler system than that (although, there are other nuances, like
>not taking the signal if the last theoretical trade was a winner).
>The thing about that system is that it makes logical sense. It's easy
>to understand and picture in the mind. The idea is that if a security
>(stock or commodity) breaks out of a trading range, there is likely
>to be a continuing momentum that will carry it into a long term
>trend. There are lots of false signals, but when it is not false, the
>profits are really big. Lots of people who follow the Turtle system
>have made lots of money with it over the years.
> >
> > When I read your code with no explanatory notes, I was puzzled as
>to the meaning of the lines, and frankly I still am. The coefficiet
>0.001 is likely good only for the QQQ. True?
>No, this is not true. The D-ratio=(H-L)/(H+L) has similar range for
>all stocks, see the suggested explorations or wait for the detailed
>examples later today.
>I suppose it has to be different if a $200 stock is being used,
>right?
>Negative again, the D-ratio is a relative quantity, not an absolute
>one.
>What's the fundamental basis of the 0.001?
>It is a coefficient to arrange D-ratio values in the scale 0-100
>The H+L? The other parameter coefficients? If the H/L spread shrinks
>at the peak in price, then adding the high to the low value and
>multiplying it by a bigger number compensates for when the H/L spread
>increases at the trough, where you multiply the sum of the H and L by
>a lower number.
>H+L is used as an "average" of intraday values instead of C.
>In MANY indicators we use (H+L)/2 or (H+L+C)/3 or (H+L+2*C)/4 instead
>of C.
>You buy when the range is greater than the product of H+L*0.001*43.
>It's all completely and totally empirical.
>I do not see any problem to use empirical relations. If, someday,
>they will be replaced by more "scientific" ones, I will be glad, but
>they will be much more complicated and there is no doubt about it.
>14 and 43 were the optimal values for QQQ, nothing more, nothing
>less. I did not speak for something universal, it was just an
>individual study.
>Again, as I said for the 0.001 coefficient, I suspect each stock
>would have to have its own d1 and d2 parameter values, right? If
>that's true, it would be too complicated, at least for my feeble
>mind, to keep track of all those different parameter values.
>I agree with you but what can I do ?
> >
> > >>Do you understand the same question for the indicators you use ?
>Do you "see" the meaning of StochD()<30 for example?>>
> >
> > This may come as a surprise to you, Dimitris, but I try to avoid
>use of indicators as much as possible, and most especially if I don't
>understand the indicator. I like ADX as a trigger for the beginning
>of a trend. I can go into a long explanation as to why, but that
>would be counterproductive on this board. I use simple trailing
>stops. I don't use Stochastics, RSI, DEMA, TEMA, etc., etc.
>I consider the analytic RSI formula by far more complicated than the
>H/L relations. As for DEMA and TEMA, I will not discuss it at all.
>If you really have a "common" understanding for TEMA, please, do not
>hesitate to post it, everybody will read it.
>All I was saying was when you suggest and post a trading system on
>the board, it would be nice to have a short explanatory dialogue for
>its use or in support of its basis. That's all. You said Stoch()<30
>is much more complicated than a simple H/L relation. You are right.
>But, I didn't understand the particular H/L relation you were
>proposing. Thanks for your contribution.
>I will post the D-ratio study today. After reading it, please ask
>anything you want and I will try to respond.
>DT
>PS: Shall we open a new thread re:Turtles ?
>Do you Have any experience ? Any recent backtesting results ?
>Have you coded the system ?
> >
> > AV
> >
> >
> > ----- Original Message -----
> > From: dtsokakis
> > To: amibroker@xxxx
> > Sent: Sunday, June 23, 2002 4:06 PM
> > Subject: [amibroker] Re: QQQ Individual Analysis
> >
> >
> > Well,
> > I try to respond as fast as possible, as you see.
> > Different questions from different points of view.
> > It is not only QQQ, we have had the same discussion with Don
>today
> > for MSFT.
> > QQQ was an example. Just an example.
> > This excellent forum is strange.
> > When I give Trade the Market systems, many people ask for
>individual
> > analysis. When i give EXACTLY individual analysis, you ask "why
>use
> > it ONLY on the QQQ? ".
> > It is strange, isnt it?
> > Many people ask for a "copy/paste" formula without explanations.
> > When I give a SIMPLE, PRE-INDICATOR formula with H and L only,
>[have
> > you seen anything more simple than this ?], nobody is satisfied,
> > because they want to learn something "behind" the formula.
> > Please, do not make simple things to look as complicated.
> > When the stock is near the peak the H/L spred shrinks, when it is
> > near the trough the spread widens and thats all.
> > It is just an observation and it may be translated to some
>trading
> > rules. Is it more complicated than RSI mechanism ?
> > I think no.
> > I would like to read what you do not understand, but, before
>posting
> > your [interesting] question just think : Do you understand the
>same
> > question for the indicators you use ?
> > Do you "see" the meaning of StochD()<30 for example ?
> > I think it is much more complicated than a simple H/L relation ?
> > I would like to have your opinion.
> > DT
> > --- In amibroker@xxxx, "Avcinci" wrote:
> > > DT,
> > >
> > > Several people have asked you to explain the rationale behind
>your
> > latest "system", but you have not responded. I think it would be
>very
> > helpful that, whenever someone submits a trading idea to the
>group,
> > he/she at least includes a short paragraph with some brief
> > explanation or support for the rationale behind the system or
>idea. I
> > look at the code and it baffles me what it is trying to say. You
>buy
> > when the today's range exceeds 0.1% of a fudge factor multiplying
>the
> > sum of the high and low, and you sell when today's range is less
>than
> > this product. What the heck does that mean? And, why use it ONLY
>on
> > the QQQ? Very baffling.
> > >
> > > Al V.
> > > ----- Original Message -----
> > > From: Nurudin Kaba
> > > To: amibroker@xxxx
> > > Sent: Sunday, June 23, 2002 6:18 PM
> > > Subject: RE: [amibroker] Re: QQQ Individual Analysis
> > >
> > >
> > > 14,43 happen to be the optimum numbers...though it does form
>an
> > island using Herman's 3D graphing via excel...
> > >
> > > Interesting...but still don't understand the rational behind
>the
> > system.
> > >
> > > Thanks
> > > -----Original Message-----
> > > From: dtsokakis [mailto:TSOKAKIS@xxxx]
> > > Sent: Sunday, June 23, 2002 12:10 PM
> > > To: amibroker@xxxx
> > > Subject: [amibroker] Re: QQQ Individual Analysis
> > >
> > >
> > > As you see in settings, the final result is for Long &
>short
> > trades.
> > > If you select Long only, or Short only, you get the
>respective
> > results
> > > ///QQQ [14,43]=+177% [LONG]
> > > ///QQQ [14,43]=+359% [SHORT] AND
> > > ///QQQ [14,43]=+1175% [LONG & SHORT]
> > > for parameters values 14, 43
> > > DT
> > > --- In amibroker@xxxx, "Nurudin Kaba" wrote:
> > > > DT, what is the premise behind HL.afl.
> > > >
> > > > Thanks
> > > > -----Original Message-----
> > > > From: dtsokakis [mailto:TSOKAKIS@xxxx]
> > > > Sent: Sunday, June 23, 2002 11:16 AM
> > > > To: amibroker@xxxx
> > > > Subject: [amibroker] Re: QQQ Individual Analysis
> > > >
> > > >
> > > > Peter,
> > > > My data begin on 3/1/2000.
> > > > See analytic #
> > > > http://groups.yahoo.com/group/amibroker/message/20095
> > > > DT
> > > > --- In amibroker@xxxx, "bluesinvestor"
> > wrote:
> > > > > Dimitris,
> > > > >
> > > > > Again great performance and yet not the same trades
>or
> > > results ...
> > > > does your
> > > > > history go back further than Jan 2000?
> > > > >
> > > > > Peter
> > > > > -----Original Message-----
> > > > > From: Dimitris Tsokakis [mailto:TSOKAKIS@xxxx]
> > > > > Sent: Sunday, June 23, 2002 11:59 AM
> > > > > To: amibroker@xxxx
> > > > > Subject: [amibroker] QQQ Individual Analysis
> > > > >
> > > > >
> > > > > If you are fed up from my Composite Analysis, let
>us
> > make a
> > > break
> > > > with an
> > > > > individual one.
> > > > > Simple logic, simple profits, pre-indicator era.
>[High,
> > Low
> > > and
> > > > thats all
> > > > > !!]
> > > > > Dimitris Tsokakis
> > > > >
> > > > >
> > > > > Yahoo! Groups Sponsor
> > > > > ADVERTISEMENT
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > Your use of Yahoo! Groups is subject to the Yahoo!
> > Terms of
> > > > Service.
> > > >
> > > >
> > > > Your use of Yahoo! Groups is subject to the Yahoo!
>Terms of
> > > Service.
> > >
> > >
> > > Your use of Yahoo! Groups is subject to the Yahoo! Terms of
> > Service.
> > >
> > >
> > > Your use of Yahoo! Groups is subject to the Yahoo! Terms of
> > Service.
> >
> >
> > Yahoo! Groups Sponsor
> > ADVERTISEMENT
> >
> >
> >
> >
> > Your use of Yahoo! Groups is subject to the Yahoo! Terms of
>Service.
>
Chat with friends online, try MSN Messenger: Click Here
|