|
PureBytes Links
Trading Reference Links
|
Blitz,
To complete my previous post, an explanation of the
"Bouncing" aspect of "Bouncing Ticks" is in order.
Assume we have a bar with a high of 100 and a low of
90. With Bouncing Ticks set to 10% (the default
setting), TS will bounce back 1 point to test for an
exit condition before it then proceeds to make it's
assumptions about the path from O - H/L - C. BT moves
as a % of the range of the bar, not an absolute %.
I keep Bouncing Ticks set to 0%, because I believe
that any exit triggered as a result of BT is faulty
information. The chances are pretty steep that over a
long EOD test period (years) the market will always
"bounce back" by a certain % of every bar's range to
test your stop. Also, if your stop is getting hit by
BT it's probably set to tight to be realistic anyway
and the only way you could show good results would be
to couple that tight stop with the %Trailing stop. Of
course, then you're really creating a work of fiction.
You can reliably test a VolBreak type system on EOD
data though, if you set BT to 0% and exit on the
close, or tomorrows open, or the following close, etc.
Then if you find some result you want to test further,
you've got to load up an intraday chart, build
"synthetic" daily bars, and test your entry with a
more sophisticated exit.
Lance
|