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

[amibroker] Re: Backtest using equity curve



PureBytes Links

Trading Reference Links

My comments are below.

--- In amibroker@xxxxxxxxxxxxxxx, "Howard Bandy" <howardbandy@xxxx>
wrote:
> If
> we have designed trading systems based on recognition of patterns
that
> precede profitable trading opportunities, and if those patterns are
> persistent, then we no longer have random, independent outcomes.  

I think you're confusing randomness, independence and stationarity.

>Our
> trading systems do have serial dependencies and upward sloping
equity
> curves.  

An upward sloping equity curve has *absolutely nothing to do with
serial dependencies*.  And further, not all systems have them.  Which
statistical tests are you using, BTW?

 
> 
> Howard
> 
> 
> 
> -----Original Message-----
> From: quanttrader714 [mailto:quanttrader714@x...]
> Sent: Saturday, November 15, 2003 8:30 PM
> To: amibroker@xxxxxxxxxxxxxxx
> Subject: [amibroker] Re: Backtest using equity curve
> 
> 
> 
> Well, when you figure out how to do that would you please post your
> methodology so I can give it a spin?  I'll keep an open mind but my
> initial testing wasn't promising.  And logically, it seems to me
that
> equity feedback is just an application of the gambler's fallacy
unless
> serial dependencies exist between your trades which is not the case
> for mine.
> 
> --- In amibroker@xxxxxxxxxxxxxxx, "Herman vandenBergen"
<psytek@xxxx>
> wrote:
> > After reading my post below i realize all this can be tested in
the
> new
> > Portfolio tester using PositionScore, no longer a need to write
> separate
> > code as we used to do :-) must try this soon.
> >
> > Thanks Tomasz! There are always numerous new applications made
> possible with
> > the new features you introduce...
> >
> > herman.
> >   -----Original Message-----
> >   From: Herman vandenBergen [mailto:psytek@x...]
> >   Sent: November 16, 2003 9:49 AM
> >   To: amibroker@xxxxxxxxxxxxxxx
> >   Subject: RE: [amibroker] Re: Backtest using equity curve
> >
> >
> >   imho Equity feedback can only be expected to work in short term
> trading
> > systems, i.e. something like 1 to 5 day trades.
> >
> >   It would make sense to use separate equity curves for Long and
> Short
> > positions. The L&S composite equity curve may go up while the
Short
> equity
> > curve is going down, same for the Long side. The feedback equity
> array
> > should be sampled with the last trade of the same type, i.e. do
not
> include
> > the last Long trade period in the evaluation of the Short equity,
> this is
> > essentially a Cash priod for the Short equity and carries no
> information -
> > unless your system is perfectly symmetrical...
> >
> >   herman.
> >     -----Original Message-----
> >     From: quanttrader714 [mailto:quanttrader714@x...]
> >     Sent: November 16, 2003 9:38 AM
> >     To: amibroker@xxxxxxxxxxxxxxx
> >     Subject: [amibroker] Re: Backtest using equity curve
> >
> >
> >     I took a quick look at this a while ago and wound up with
> similar
> >     findings.  For the (good) systems I looked at, equity feedback
> (only
> >     taking trades when equity is above MA) hurt performance. 
Then I
> >     looked at what would happen if only taking trades when equity
is
> below
> >     the MA and the results were even worse.  At which point I
> dropped the
> >     study and didn't bother to look at mediocre systems or try to
> confirm
> >     the results with another method.  Not saying this is the
> definitive
> >     answer, but would love to see quantitative evidence to the
> contrary.
> >
> >     --- In amibroker@xxxxxxxxxxxxxxx, "giggollo99" <giggollo@xxxx>
> wrote:
> >     > Hi Herman, Glenn:
> >     >
> >     > I did some studies on this idea about a year ago with
futures
> data.
> >     > For example, as Glenn suggested, "No new entries if a closed
> trade
> >     > crosses below a moving average of the equity curve and
> re-enter
> >     when
> >     > a closed trade crosses above the moving average."
> Interstingly I
> >     > found one potential problem to be that you would miss out on
> the
> >     > winning trades that were carrying the equity back over its
MA
> and
> >     > would trigger you to start taking your system trades again.
> Without
> >     > equity feedback, those trades add to your bottomline and
make
> a big
> >     > difference i found. However, you are always taking the
trades
> that
> >     > take you back under the equity curve. This is certainly an
> >     > interesting approach and requires further study perhaps with
> other
> >     > data sets.
> >     >
> >     > Tomasz: Thanks a lot for the portfolio backtester..its
> >     amazing..look
> >     > forward to seeing the pyramiding feature implemented in the
> future
> >     > hopefully
> >     >
> >     > best regards to all
> >     > g
> >     >
> >     >
> >     >
> >     >
> >     > using  and found that one potential problem with using
equity
> curve
> >     > to start/stop your system
> >     >
> >     >
> >     >
> >     > --- In amibroker@xxxxxxxxxxxxxxx, "Herman vandenBergen"
> >     > <psytek@xxxx> wrote:
> >     > > Metastock is a few years behind me :-)
> >     > >
> >     > > Yes it is great fun to create combos; systems build up
from
> >     > several systems.
> >     > > You can use the signals from a slow system to qualify
> signals
> >     from
> >     > a fast
> >     > > system (like a trend qualifier). Or you can use the Long
> equity
> >     > from slow
> >     > > system one as a trend indicator for system number two. You
> can
> >     > even pack a
> >     > > large number of systems (perhaps including some duplicate
> systems
> >     > but with
> >     > > different parameters) into one piece of code and optimize
> for the
> >     > best
> >     > > signal combination from the lot of them. When you separate
> Long
> >     > and Short
> >     > > equities they can become trend indicators for the system
> under
> >     > test, it
> >     > > uniquely reflects the trend sensitivities for the system,
> much
> >     > better then a
> >     > > totally unrelated trend indicator... imho
> >     > >
> >     > > Unlimited possibilities, aren't you glad you bought
> AmiBroker :-)
> >     > >
> >     > > herman
> >     > >   -----Original Message-----
> >     > >   From: Glenn [mailto:glennokb@x...]
> >     > >   Sent: November 12, 2003 3:10 PM
> >     > >   To: amibroker@xxxxxxxxxxxxxxx
> >     > >   Subject: [amibroker] Re: Backtest using equity curve
> >     > >
> >     > >
> >     > >   Hi Herman,
> >     > >
> >     > >   Thanks for your reply. Excellent!
> >     > >
> >     > >   Have you done much testing with your equity curve to
> change your
> >     > >   systems parameters?
> >     > >
> >     > >   I've done most of my backtesting with Metastock and
> Tradesim
> >     and
> >     > been
> >     > >   waiting for ages for Tradesim to have this in it.
> >     > >
> >     > >   Ami has been able to do it all along!
> >     > >
> >     > >   Cheers Glenn
> >     > >
> >     > >   --- In amibroker@xxxxxxxxxxxxxxx, "Herman vandenBergen"
> >     > <psytek@xxxx>
> >     > >   wrote:
> >     > >   > You can cascade as many systems as you like, even
> different
> >     > ones,
> >     > >   and use
> >     > >   > the Equity from the previous one as a parameter in the
> next
> >     > system.
> >     > >   I know
> >     > >   > this can be done with the old backtester and think it
> should
> >     > also
> >     > >   work in
> >     > >   > the new PF tester.
> >     > >   >
> >     > >   > // system one code here
> >     > >   > E1 = Equity(1);
> >     > >   >
> >     > >   > // System two code here
> >     > >   > Buy = Buy and (some function of E1);
> >     > >   > E2 = Equity(1);
> >     > >   >
> >     > >   > // System three code here
> >     > >   > Buy = Buy and (some function of E2);
> >     > >   > E3 = Equity(1);
> >     > >   > etc.
> >     > >   >
> >     > >   > You essentially redefine the buy signal as often as
> needed
> >     (afl
> >     > >   executes
> >     > >   > line after line and never looks back), the last
> definition
> >     > will be
> >     > >   what
> >     > >   > determines your results.
> >     > >   >
> >     > >   > Herman
> >     > >   >
> >     > >   >   -----Original Message-----
> >     > >   >   From: Glenn [mailto:glennokb@x...]
> >     > >   >   Sent: November 12, 2003 1:35 PM
> >     > >   >   To: amibroker@xxxxxxxxxxxxxxx
> >     > >   >   Subject: [amibroker] Backtest using equity curve
> >     > >   >
> >     > >   >
> >     > >   >   Hi,
> >     > >   >
> >     > >   >   I'm was wondering if it is possible in AB to
> incorporate the
> >     > >   equity curve
> >     > >   >   of a system within a backtest, using it to test the
> >     > following:
> >     > >   >
> >     > >   >   a. No new entries if a closed trade crosses below a
> moving
> >     > >   average of
> >     > >   >   the equity curve and re-enter when a closed trade
> crosses
> >     > above
> >     > >   the
> >     > >   >   moving average. Another idea is to use a percentage
> on the
> >     > equity
> >     > >   curve
> >     > >   >   instead of a moving average.
> >     > >   >
> >     > >   >   b. Using the above also test tightening the actual
> trailing
> >     > stop
> >     > >   on the
> >     > >   >   open trades. ie: if a closed trade crosses below a
> moving
> >     > average
> >     > >   (or
> >     > >   >   whatever) then instead of using a 3 x ATR stop then
> use a 2
> >     > x ATR
> >     > >   stop
> >     > >   >   on the open trades.
> >     > >   >
> >     > >   >   Note that the trades in between the exit and entry
> need to
> >     be
> >     > >   tracked for
> >     > >   >   the re-entry.
> >     > >   >
> >     > >   >   If this is possible, do you know how to set it up
> please?
> >     > >   >
> >     > >   >   Cheers, Glenn
> >     > >
> >     > >
> >     > >         Yahoo! Groups Sponsor
> >     > >               ADVERTISEMENT
> >     > >
> >     > >
> >     > >
> >     > >
> >     > >   Send BUG REPORTS to bugs@xxxx
> >     > >   Send SUGGESTIONS to suggest@xxxx
> >     > >   -----------------------------------------
> >     > >   Post AmiQuote-related messages ONLY to:
> amiquote@xxxxxxxxxxxxxxx
> >     > >   (Web page: http://groups.yahoo.com/gro
> up/amiquote/messages/)
> >     > >   --------------------------------------------
> >     > >   Check group FAQ at:
> >     > > http://groups.yahoo.com/group/amibroke
r/files/groupfaq.html
> >     > >
> >     > >   Your use of Yahoo! Groups is subject to the Yahoo! Terms
> of
> >     > Service.
> >
> >
> >
> >     Send BUG REPORTS to bugs@xxxx
> >     Send SUGGESTIONS to suggest@xxxx
> >     -----------------------------------------
> >     Post AmiQuote-related messages ONLY to: am
iquote@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 the Yahoo! Terms of
> Service.
> >
> >
> >         Yahoo! Groups Sponsor
> >               ADVERTISEMENT
> >
> >
> >
> >
> >   Send BUG REPORTS to bugs@xxxx
> >   Send SUGGESTIONS to suggest@xxxx
> >   -----------------------------------------
> >   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 the Yahoo! Terms of
> Service.
> 
> 
> 
> 
> 
> 
> Yahoo! Groups Sponsor
> 
> 
> 
>
<http://rd.yahoo.com/SIG=12c803vs6/M=259395.361467
4.4902533.1261774/D=egroup
>
web/S=1705632198:HM/EXP=1069039790/A=1524963/R=0/*
http:/hits.411web.com/cgi-
> bin/autoredir?camp=556&lineid=3614674¢ç=egroupweb&pos=HM>
> 
> 
> 
>
<http://us.adserver.yahoo.com/l?M=259395.3614674.4
902533.1261774/D=egroupmai
> l/S=:HM/A=1524963/rand=725798141>
> 
> 
> Send BUG REPORTS to bugs@xxxx
> Send SUGGESTIONS to suggest@xxxx
> -----------------------------------------
> 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 the Yahoo!
> <http://docs.yahoo.com/info/terms/>  Terms of Service.


------------------------ Yahoo! Groups Sponsor ---------------------~-->
Buy Ink Cartridges or Refill Kits for your HP, Epson, Canon or Lexmark
Printer at MyInks.com. Free s/h on orders $50 or more to the US & Canada.
http://www.c1tracking.com/l.asp?cid=5511
http://us.click.yahoo.com/mOAaAA/3exGAA/qnsNAA/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/