|
PureBytes Links
Trading Reference Links
|
Al,
I can't answer your questions. I'm not versed in CAR, my drawdowns
are usually limited due to using fairly tight moving stops, and I
can't envision why the percent volatility model would only work for
longer term systems. My range was only a year and showed good
results. However, you may have noticed my position scoring was based
on volatility (C/ATR) (and chosen from a universe of high beta
stocks) so that may invalidate the results for a system based on
other scoring methods (?).
In the last chapter of Tharp's book he shows that a % volatility
model produces a return of almost 10X over an equal equity model
(over $2M compared to just $231K). The period was 5.5 years and the
system was the same -- only the position scoring was different! I
suspect he had many more trades in the model that did better but
that's okay -- if necessary, I'd trade 10 times as much for 10X
returns.
If you have access to his book, I suggest you read chapter 12 and
then explain the different sizing methods to me. Right now, I'm
afraid the subtleties are whizzing right over my head.
However, if the differences are as significant as Tharp contends,
it's important we understand this.
Dan
--- In amibroker@xxxxxxxxxxxxxxx, Al Venosa <advenosa@xxxx> wrote:
> danielwardadams wrote:
>
> > Just FYI, I got some pretty good results backtesting with the
> > following:
> >
> > Capital = -20 ; //20% of Equity.
> > RiskCapital = .05*Capital ;
> > NumberOfShares = RiskCapital/(2.37*ATR(10)) ;
> > PositionSize = Max(-20,NumberOfShares * BuyPrice [i]) ;
> >
> > I think this achieves what we initially wanted (volatility based
> > sizing with diversification of at least 5 stocks). I think there
is
> > some subtle difference between this and what Ed proposed for
> > PositionSize yesterday but I'm not sure what it is (?).
> >
> Dan:
>
> This is a good idea. However, I tried testing it systematically
relative
> to the simple 1% volatility model tonight, and I got the following
> results. First, using the simple code PositionSize = -1*C/Stoploss
and
> using SetOption("MaxOpenPositions",15), I can get all equity used
up in
> all trades whether or not I use 50% margin. Often, there would be
as
> many as 12 or 13 trades open at one time. So, setting the maximum
no. of
> open positions to a large number like 15 solves the inability to
use all
> equity problem. You don't need Ed's Max part of the positionsize
> statement. And, if you use 50% margin, you also use up all equity
and
> your profit (and MDD) also go up substantially, as expected. Using
the
> equal equity model (posqty = 5, positionsize = -100/posqty), I get
> higher CAR than the 1% model but also slightly higher MDD. Using
margin,
> however, the CAR triples while the MDD increases by over 50%. Using
your
> approach with no margin, I get comparable results as with my
approach
> above. However, strangely, not all capital is used in each trade.
Using
> margin, a lot of capital is left unfunded. Bottom line: it's
difficult
> to beat the equal equity model. The % volatility model simply
doesn't
> perform as well. Again, I ask could this be due to the fact that
> volatility doesn't have that much effect with short term systems?
>
> Al Venosa
------------------------ Yahoo! Groups Sponsor --------------------~-->
Make a clean sweep of pop-up ads. Yahoo! Companion Toolbar.
Now with Pop-Up Blocker. Get it for free!
http://us.click.yahoo.com/L5YrjA/eSIIAA/yQLSAA/GHeqlB/TM
--------------------------------------------------------------------~->
Check AmiBroker web page at:
http://www.amibroker.com/
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/
|