Stop Loss hunter strategy

115 views
Skip to first unread message

Javier

unread,
Aug 23, 2011, 2:36:33 PM8/23/11
to JBookTrader
When a lot of stop loss orders are executed almost at the same level,
a important market drop comes up, do you think that is possible to
catch some of this drop in advance? Which JBT indicators should we use?

Judson Wilson

unread,
Aug 23, 2011, 2:56:54 PM8/23/11
to jbook...@googlegroups.com
I would expect a sharp change in book balance, price, (with a specific correlation in direction) and a volume spike.


--
You received this message because you are subscribed to the Google Groups "JBookTrader" group.
To post to this group, send email to jbook...@googlegroups.com.
To unsubscribe from this group, send email to jbooktrader...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/jbooktrader?hl=en.


nonlinear

unread,
Aug 23, 2011, 7:08:12 PM8/23/11
to jbook...@googlegroups.com

You may want to experiment with the VolumeAcceleration indicator which is checked in. I found that it is beneficial to have it as part of your strategy. My theory for using it is this: the move down will continue until it triggers the stops. As the stops trigger, the volume will accelerate. When there are no more stops to trigger, the volume will decelerate. That would be a good point for long entry. So, in terms of a JBT strategy, you'd go long when VolumeAcceleration is below some negative value.  In reality, things are not that simple, of course.

Javier

unread,
Sep 4, 2011, 12:49:05 PM9/4/11
to JBookTrader
> You may want to experiment with the VolumeAcceleration indicator which is
> checked in.

I've been experimenting with Pendulum based strategies and I've found
out that they are not performing well in periods like past August. I
also have tried optimizing these strategies using only hi-volatility
periods but then when the volatility is low again they can add
important losses.

I'm thinking in adding some kind of market parameters (max volatility,
max dd, ...) based on the optimized period and compare with the real
time market, if the real market data is outside these parameters then
switch off the strategy. Any thoughts about this idea?
Reply all
Reply to author
Forward
0 new messages