|
PureBytes Links
Trading Reference Links
|
OK Chuck,
1st... the web address of the Holy Grail site please. (even though I
vaughly remember a Holy Grail forum, I looked in Yahoo Groups and did
not see anything resembling a Trading Group called Holy Grail). I'll
do a little research there before embarassing myself further here. And
BTW... do you ever sleep?
And regarding my system(s) you mentioned... they are still oxymoron's
to me since they had such terrible drawdowns in the early 90's only to
survive and produce the 'eye-popping' results they did after '98. I'm
slow though, so I'll play with them awhile longer before hollering
'uncle'.
And FWIW... I remember much of your past posts about backtesting
philosophy, and happen to agree with you.
<let me ask you to tell me the price of a high-priced stock in early
2000? You may say something like $200.>
Your point is of course that in early 2000 there was a bubble that
only a few of us recognized... and that we used as a signal to take
our profits and exit the market.
Anyway... I'm through for the night (I can't do this 24 hrs a day like
you).
Tomorrow...
Phsst
--- In amibroker@xxxxxxxxxxxxxxx, Chuck Rademacher
<chuck_rademacher@x> wrote:
> Phsst said:
>
> I've read this post more than once and on more than one occasion
decided to ignore it since I have not had anything to contribute to
your query.
>
> But I keep coming back to it because it strikes me as a curiousity.
>
> Like you indicated, in the past I've filtered based upon price
range, (particularly on Shorts), but I've not yet found a situation
where it paid to filter on price. And now that you have enlightned me
about split-adjusted prices of the past, I've even stopped filtering
> adjusted Short prices, preferring to use raw data filters instead.
>
> If it doesn't compromise your fiduciary responsibilities, could you
elaborate on the significance of price filtering in your backtested
trading systems?
>
>
> My reply:
>
> Thanks for replying and for giving me another chance to spell out
what I'm trying to achieve. You should see the answers to this
posting that I received at the Holy Grail site. Yikes!!
>
> I have a system, not too unlike one that you and I previously
discussed. Amongst all the other things that it is doing, it is
filtering such that it is only interested in stocks between $3 and $10.
>
> Why does it want stocks in this range? I have a few theories, but
the important thing is that if I give it any other range, it doesn't
do as well.
>
> So, my question was "Does the system want low-priced stocks or does
it want stocks very specifically in the $3 to $10 range?".
>
> Can you see the subtle difference there? In order to better
demonstrate what I'm talking about, let me ask you to tell me the
price of a high-priced stock in early 2000? You may say something
like $200. If I ask you to tell me the price of a high-priced stock
last week, you might say something like $80. Why the difference?
Because the whole market is priced differently today than it was in
early 2000.
>
> So, back to my original question. Does my system want to trade
cheap stocks or stocks between $3 and $10. As it turns out, the
system performed better in early 2000 if I raise that range to $8 to
$15. So, this would suggest to me that the system wants to trade
cheap stocks, adjusted dynamically to overall market pricing.
>
> In case it hasn't come through in everything that I've ever talked
about, I want my systems to be presented with all stocks in all
timeframes and I want my systems to make the necessary adjustments so
that they will trade well under any conditions.
>
> As you know, I adjust my volume/turnover filters according to the
average volume/turnover of the market over time. If I have a price
range filter, why shouldn't that be adjusted dynamically too? If it
should be adjusted, how should it be adjusted? I was asking for
ideas on how to make such an adjustment.
>
> While waiting for replies, I tried over 80 different methods of
adjusting the price range. The best method so far is to adjust it
based on the standard deviation from the average price. I tested this
on all active and de-listed stocks back to 1960.
>
> By making this adjustment, the system in question has:
>
> 1. A smoother equity line.
> 2. More consistent annual profit (all years with similar returns).
> 3. Higher CAR and lower MaxDD.
>
>
> Is it more curve fit or is it adjusting to market dynamics? I think
it is the later and I realise that there are those who will think it
is the former.
>
> The logic is staying in!!
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
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/
|