|
PureBytes Links
Trading Reference Links
|
First, I cant remember how Behold! is handling an indicator after the
rollover. But I would guess that it does it correct and recalculates the
indicator with the new data.
Second, what I really wanted to say, is that the real thing should be doing
a rollover and using real data. Wouldnt that be perfect, wasnt SystemWriter
doing that? It just makes me thinking, that if I use the same contract but
use differnet method to combine them that the result is so different. But it
seems instead of asking for the most logical solution, everybody is arguing
for the best work arround.
Volker Knapp
Wealth-Lab Inc.
http://www.wealth-lab.com
http://www.wealth-lab.de
++-----Ursprungliche Nachricht-----
++Von: DH [mailto:catapult@xxxxxxxxxxxxxxxxxx]
++Gesendet: Samstag, 13. April 2002 02:37
++An: Omega List
++Betreff: Re: AW: Why Continous contract software
++
++
++Yeah, what Gary said. :-)
++
++> Cont. contracts or back adjusted contracts are a real turn
++off, I think.
++
++Whatever floats your boat. I use them for system development and
++testing. I wouldn't enjoy working without them.
++
++> Stridsman points out the risk of using the "wrong" kind of
++contract oout in
++> his book.
++
++Stridesman makes several logic errors in his book. I've debated him
++publicly on the matter but he "just doesn't get it." Obviously, he has a
++firm grasp of the obvious when he says doing things wrong is wrong. My
++answer to that.... 'well duh.' :-) Slightly more serious answer, read
++Bob Fulks' article. It's short but his conclusions are spot on. It
++covers the subject much more completely than any book I've seen.
++
++>I always felt there is only one way to really test systems and
++> that is with the real contract.
++
++There is always more than one way to do things. Any trader should know
++that.
++
++Bottom line, your backtesting method should give trades at exactly the
++same date and time as they would have occurred in real trading. Your
++profits and losses should be identical to those that would have occurred
++realtime given your money management strategy. My favorite for
++backtesting is to risk an equal dollar amount on each historical trade.
++It is quite easy (well maybe not so easy - it does involve a few lines
++of code) to do that using Tradestation and continuous contracts. I'm
++sure it's available in the archives somewhere. I really don't have the
++energy to go through the whole thing again.
++
++Stridesman claims this cannot be done in Tradestation. After being shown
++HOW it can be done in Tradestation, he still claims it cannot be done
++and you must use Excel. It sells more books I guess. Judge his
++credibility for yourself.
++
++--
++ Dennis
++
|