Discussing the strategy VECTORIAL DAX (M5)
09/11/2020 at 8:25 AM #14391509/11/2020 at 8:57 AM #14391709/11/2020 at 4:28 PM #14395109/11/2020 at 5:40 PM #143953
good question, the answer can go many ways!
For the dax I first determined 2% stoploss and timezone 8-22h & no weekend position but have overnight and to test with tradetype=1 with re-entry on and always no MM. If using a 2% stoploss, just use the trailing stop at 0.5% or 1% but don’t optimise other variables. Put all other exit criteria off.
Focus on the settings of vectorial and the angles.
Then finetine with stochastic rsi indicator and perhaps the mr indicator.
On the other hand perhaps it’s better to have a long & short strategy separated. It’s something to test.09/11/2020 at 9:30 PM #143974
Thanks. That is almost what I did.
It’s OMX I’m trying on, but my result was horrible.
I’ll give it another go and start optimizing from the back instead. Angle first, then vectorial. Or are you doing everything at once? Got too many combinations last time, so had to split it up.09/11/2020 at 10:35 PM #143979
I’ve been working on the dax, now on long only with 1% stoploss and no weekend position. Will take it look at omx too.
Try not everything at once and focus on long only, 1% stoploss & reentry on. Just take 25 as angle value, then optimise the vectorial averages. If something pops up, not necessary the highest payout, maybe more trades is better, narrow it down the range for optimising and add now only 1 angle and try again. Actually results should be good without the rsi.
Long is a bit easier. Here’s what I got on the dax and to show what’s it’s like with mr indicator on/off
btw check the flag numbers. Made a mistake in the end.
1 user thanked author for this post.09/11/2020 at 11:43 PM #143985
oke that was a bit messed up above, but here’s the file for the dow long only with updated mr, in fact I’ve added another one to see the difference using hours between 0-9 and/or with 8-9.
09/12/2020 at 9:53 AM #14400309/12/2020 at 10:01 AM #14400409/13/2020 at 2:49 PM #144096
@paisantrader, I played around a bit with the sys-05 vectorial -us100 v5.(1) code on OMX and got interesting backtest results. It might not be very robust though and honestly I don’t have a deep enough understanding of how the trailing stop and much of the code works. Mostly just optimized angle, nbChandelier and TP, SL…
09/13/2020 at 3:53 PM #14410709/13/2020 at 4:02 PM #14410909/13/2020 at 5:35 PM #14413009/13/2020 at 6:19 PM #14413809/13/2020 at 6:26 PM #144139
Thanks. I am certain there’s lots of room for improvement and optimization.
BTW not sure I understood what part was wrong with the %TrailingStop – does it matter in these versions or does it only matter if you use breakeven?