|
PureBytes Links
Trading Reference Links
|
Hello Adrian,
Looking at your gif I see you marked turning points at some swinghighs
and swinglows while other valid swinghigh/swinglows were overlooked.
Mechanically trading with swings there is no way of avoiding the false
signals that crop up in the middle of the nice moves, if you make your
swing look-back too far you miss most of the move.
Percent Profitable Trades using swing systems is not usually very high
because of the number of fake-out swing trades.
I personally like a system that has a high number of Percent Profitable
trades and suggest you look at something other than swings.
A starting point may be to look at the system MACount by Robert
Linders that was posted on this list on Oct. 26th
As far as stops go here is my opinion
1. End of day Exit ( May improve results, easily tested. )
2. Profit Target ( Stops you out with a profit )
3. Trailing Stop ( Occasionally stops with a profit )
4, Money Management Stop ( Disaster Protection )
regards
foolsgold
AP> Hi,
AP> I'm wondering if someone could give me some suggestions here. I'm
AP> trying to work out how to program a particular type of trailing stop.
AP> Typically many use the Donchian X day trailing stop. I am trying a
AP> variation. I have attached a graph showing what I am trying to achieve.
AP> Basically I want to find the current price extreme, say a high, then
AP> trail a stop below that high X bars back from the low of the high bar,
AP> calculated via the low I.e. I move back from the low of the high bar
AP> until I hit X new price lows and that becomes the stop point. This may
AP> involve going back many bars before being able to find those X news
AP> lows, or highs in the case of a low pivot. The example shown works on
AP> X=2. What this will means is that when a small choppy consolidation
AP> forms it wont necessarily stop you out as would a Donchian stop.
AP> I know I probably need to use the SwingHigh/Swinglow functions, but
AP> because of the complexity of not knowing how many bars may be involved
AP> it gets complicated. I was hoping someone here could provide some
AP> suggestions on how to tackle this. It would much appreciated.
AP> Thanks,
AP> Adrian Pitt
|