|
PureBytes Links
Trading Reference Links
|
Fred,
Since we both agree that the DD's reported by AB on portfolio
backtests are 'suspect', then this discussion is academic.
You focus intensely upon DD's, and based upon what I percieve about
your preferred method of backtesting single issues then I'd agree that
DD's constitute a major consideration in trading system design where
you are exposing yourself to significant risk on any given trade.
Since I trade portfolios made up of over 8,000 plus stocks, I can have
anywhere from as little as zero to as many as 60 trade signals on any
single day. The larger the number the stronger the trend, and when I
get more than 20 signals on a single day an alarm goes off in my head
that tells me to watch closely for major changing mkt conditions.
In my method of swing trading, I only risk a small, fixed amount on
each trade. And each trade only last an average of 5 or 6 days and is
fully protected with trailing stop loss orders. In order to experience
significant DD's, I'd have to experience a series of sequential losses
on a large number of small trades on a major mkt turn, and my trading
system stops generating trades when the ADX, PDI, MDI levels expose
major changes in mkt conditions. So not only do I have trailing stop
loss protection on individual trades, I have 'trading system'
protection insofar as the system is designed to trade ONLY when the
prevailing trend is intact.
I can do a quick 'eyeball' check on the system 'portfolio' backtest
DD's by exporting the system trades to Excel... sorting them
chronologically... doing 'periodic' tallies of P/L and observing the
increase/decrease in capital.
But I have to admit that I am getting a little tired of doing the
manual 'eyeball' check in Excel and am developing my own software to
take the AB exported trades and automate the process of reporting
reliable DD info for portfolio trading systems.
What I am telling you is that your fixation on DD's is relevent to
single issue testing or portfolio testing where your capital exposure
to the mkt is *extreme* on any given trade.
But as you reduce you exposure on a per trade basis, and design your
trading systems to react to changing mkt conditions, DD's are smoothed
considerably (even though DD's should be periodically monitored manually).
So, keep your focus where DD's are involved in 'significant risk'
trades, but where you know that the reported DD's figures are not
reliable on portfolio trades, ease up a bit.
Regards,
Phsst
--- In amibroker@xxxxxxxxxxxxxxx, "Fred" <fctonetti@xxxx> wrote:
> Phsst,
>
> As I recall the results you showed had 60+% DD's. Admittedly AB
> doesn't show results of portfolio trading properly and as a result
> there was really not much to comment on was there ?
>
> --- In amibroker@xxxxxxxxxxxxxxx, "phsst" <phsst@xxxx> wrote:
> > Yuki,
> >
> > Glad to see you back from your vacation, and I hope the portion of
> > your vacation exclusive of Larry Williams was enjoyable.
> >
> > Over the years, I've been to two or three 'guru' seminars, and like
> > you, I've generally been disappointed in the quality of the
> > presentations, not to mention the presentors.
> >
> > Frankly, until I started watching the AmiBroker board, the single
> best
> > published work I had read was Linda Rasche's 'Street Smarts' book.
> > Linda did not hold much back and did not let her ego get involved in
> > the presentation at all. Her presentation primarily focused upon
> > futures, while I tend to focus upon stocks. But her approach worked
> > well in any trading environment.
> >
> > I thought I had developed a few good trading systems before getting
> > involved in this AmiBroker board. And when I recently posted my best
> > trading system results, there wasn't much feedback from others
> > regarding comparable results.
> >
> > Then I concluded that like the old Poker Player axiom... if you
> don't
> > know who the 'patsy' is at the table, then it is probably you!
> >
> > So I'll probably hold my cards a little closer to my chest for a
> while
> > until I get the 'feel of the game'.
> >
> > Anyway, glad you are back.
> >
> > Phsst
> >
> >
> >
> >
> >
> > --- In amibroker@xxxxxxxxxxxxxxx, Yuki Taga <yukitaga@xxxx> wrote:
> > > Hi DT,
> > >
> > > Tuesday, April 8, 2003, 3:38:59 PM, you wrote:
> > >
> > > DT> Yuki, Read
> http://groups.yahoo.com/group/amibroker/message/10072
> > > DT> to see the basic info for L. Williams well known indicator.
> Who
> > > DT> could imagine to put his name in ANY T/A link reproducing
> > > DT> Stochastic ?
> > >
> > > Sad, and surprising, but he did come off a bit like a glory
> seeker to
> > > me, so I guess I'm not all that surprised. He seemed to want to
> > > claim a lot of original TA stuff, and that other people copied
> him a
> > > lot.
> > >
> > > Quite frankly, I was surprised with his treatment on some
> things. He
> > > likes ADX a lot, but he seemed to not really understand how it
> works
> > > to me. He calls it a "trend integrity" indicator, and advises
> that
> > > when it makes extreme highs and then declines, the decline is
> showing
> > > you that there is no "integrity" in the embryonic new trend, and
> that
> > > it is bound to fail. Actually, this is quite wrong IMO, as the
> > > reversal in ADX at trend change is simply a "bleeding off", if you
> > > will, of the high level of trending that HAD been present. It is
> a
> > > symptom of the old trend, not yet referencing the new direction at
> > > all. After a while, if ADX does not stop falling and start
> rising of
> > > course, then the new direction is certainly suspect. But, that is
> not
> > > what Larry was talking about. He was talking simply about the
> initial
> > > fall in ADX from a high level, and that this fact makes a new
> > > direction suspect. That is not correct, IMO.
> > >
> > > It is only sometime later that the failure of ADX to stop falling
> and
> > > start rising with the new trend, that the new trend becomes
> suspect
> > > (and of course *all* new trends, especially reversals coming out
> of
> > > strong trending periods, are initially quite suspect). But a new
> > > embryonic trend is not suspect simply because ADX is bleeding off
> > > levels reached due completely to the strength of the old trend. A
> > > declining ADX is perhaps useful only to tell you to STAY OUT of a
> > > trade, IMO. Where the ADX is most useful to me is when it just
> > > crosses or just about crosses into strong trending territory -- on
> > > its way up -- signalling that there may be a train departing that
> you
> > > want to get on.
> > >
> > > Maybe I'm splitting hairs here, but I really thought he was
> trying to
> > > be a bit too "original" here.
> > >
> > > Best,
> > >
> > > Yuki
> > >
> > > mailto:yukitaga@x...
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Make Money Online Auctions! Make $500.00 or We Will Give You Thirty Dollars for Trying!
http://us.click.yahoo.com/yMx78A/fNtFAA/i5gGAA/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/
|