[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[amibroker] Re: 10 parallel portfolios



PureBytes Links

Trading Reference Links

b,

Acknowledged ...

As far as starting the tests goes there will undoubtedly be some 
automation etc. to the front end as well.  

Regarding one portfolio overwhelming another, I understand and 
sympathize but personally I look to develop individual trading models 
that can stand alone.  For personal investing this keeps things much 
simpler and IMHO simple is good.

Fred

--- In amibroker@xxxxxxxxxxxxxxx, "b519b" <b519b@xxxx> wrote:
> The Excel optimation is good. One would still have tell AB to run 
> the individual portfolios.
> 
> The setting one puts for position sizing would be not be important 
> if the portfolios shared profits. Since they do not, position size 
> would need to be constant so that a slightly more profitable system 
> does not overpower the other ones.
> 
> For example, consider this. System A is my primary system - the one 
> that make the profits. System B is my hedge system - the one that 
is 
> only included to smooth out the drawdowns of System A. System B is 
> mildly profitable on its own, but it has the remarkable ability to 
> be profitable when A enters one of its unpredictable but 
significant 
> drawdowns. 
> 
> In real life, I would periodically rebalance funds from system A to 
> B during normal times (or from B to A after a major drawdown by A). 
> But in testing, if I can not rebalance, then A will become larger 
> than B if compounding of profits is allowed. This would not be 
> significant at the beginning of the combined equity curve, but by 
> the middle and end A could be so much larger than B that it would 
> seem that B is a very poor stablizer.
> 
> b
> 
> --- In amibroker@xxxxxxxxxxxxxxx, "Fred" <fctonetti@xxxx> wrote:
> > b,
> > 
> > Regarding the negatives ...
> > 
> > 1.  I don't see any reason why position size would have to be a 
> > constant as long as assets are not shared between portfolios.
> > 
> > 2.  Excel has it's own automation tools and as should be noted 
> from 
> > other conversation regarding xxTool, there are already in the 
> works 
> > back end automation of report viewing in Excel.
> > 
> > Fred
> > 
> > --- In amibroker@xxxxxxxxxxxxxxx, "b519b" <b519b@xxxx> wrote:
> > > --- "Fred" <fctonetti@xxxx> wrote:
> > > > For each run that you set up 
> > > > for each of your individual portfolios 
> > > > you by definition today get a resulting 
> > > > account file which will have the equity 
> > > > curve numbers associated with it and if 
> > > > necessary it would be a simple enough 
> > > > task to combine multiple account files 
> > > > into a master account file that shows 
> > > > the overall equity line ...
> > > > 
> > > > Doesn't that suffice without making PT 
> > > > more complicated then it already is ?
> > >  
> > > Fred,
> > > 
> > > Simplicity is a virtue. 
> > > 
> > > Combining the equity curves of multiple independent runs would 
> > > provide PART of what I was thinking of. Perhaps, that part is 
> all 
> > > that is really needed. This is a tentative conclusion based on 
> the 
> > > following thoughts:
> > > 
> > > Positives of the Keep It Simple approach (no multiple portfolio 
> > > ability):
> > > 
> > > 1. This can be done NOW. No need to wait. A plus.
> > > 
> > > 2. One can use the full range of AFL settings for each system: 
> To 
> > > keep coding demands within reason, a multiple portfolio module 
> > > likely would require the systems to share various settings 
(such 
> as 
> > > the number of stocks, the initial capital, etc.) 
> > > 
> > > Negatives:
> > > 
> > > 1. Position size would have to be constant (no letting profits 
> > > compound). This is necessary so the curves can be combined, 
> > > otherwise, a curve with a slight advantage annual would in time 
> be 
> > > double the size of the other curve.
> > > 
> > > 2. The Simple method will be FAST in AB but SLOW in practice 
> since 
> > > much of what could be automated in AB will have to be done 
> > manually. 
> > > First, in AB each curve would require a modification of AFL 
code 
> so 
> > > one would be tied to the computer during testing. Second, 
> > additional 
> > > time is needed to exported to Excel and lined up with the 
others 
> in 
> > > Excel. 
> > > 
> > > Those listening in on the conversation may be aware of a point 
I 
> > > have missed. If so, speak up. 
> > > 
> > > Fred, I guess the issue boils down to the ease or difficulty of 
> > > coding multiple portfolios into your Portfolio Module. Would it 
> be 
> > > very complex? If so, it could be a pain to find and eliminate 
> bugs. 
> > > 
> > > b


------------------------ Yahoo! Groups Sponsor ---------------------~-->
Get A Free Psychic Reading! Your Online Answer To Life's Important Questions.
http://us.click.yahoo.com/Lj3uPC/Me7FAA/ySSFAA/GHeqlB/TM
---------------------------------------------------------------------~->

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 

Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/