|
PureBytes Links
Trading Reference Links
|
This is correct and the performance of my systems both being over 64k and being
able to back test to the tick level revealed much about the historical
performance, of my systems. They also run in real time with the same results as
they back tested with.
____________________Reply Separator____________________
Subject: Zero bars trade length myth
Author: pierre.orphelin
Date: 4/26/99 7:10 PM
One of the widest spread nonsense on this listios the assumption :
Zero bar for the average trae duration = crooked system.
Do the following experiencre and write this system
buy at open.
exitlong on close.
Now see the performance summary that is perfectly corect with a 0 bar
length for trades.
The reason is that TS calculates a bar from open to nextopen that makes the
duration of the bar.
A daily bar is not from open to close but open to next open ( ie 24 hours).
What confuse people with the 0 bar duration is that trailing stops used on
the same bar shows the 0 duration a,d false result because of the lck of
infomation of OHLC tick sequence.
TS4 uses Bouncing Ticks approximation that is not reality ( you may vary %
of bouncing ticks to fake vzriois situation,, but it's only an
approximation).
Unfair peple may use this, but it does not mean thta all 0 bars systems
are false.
Any qyqtem that do not makes use of $ trailing stop and produces exits on
every bar( close or xwhatever you want known at open) is not suspect.
For more info, read the manuals, previous posts from me one year ago, or
upgrade to TS2000i and use tick data to make real simulations ( in this
case TS2000 backtests to the tick level)
Sincerely
-Pierre Orphelin
Représentant exclusif de Omega Research en France.
web: http://www.sirtrade.com
|