I got your PM and try to answer but your PM is disable...
Maybe because you don't have any posts
J.
No the user, had changed a setting, not allowing others to send him PM, I have changed this, so you can try again..
The ability sending PM's here is not depending form number of postings a user made.
Yes, Jforex is the best solution for muti-pair strategies. Has been 2 years now.
After 1000s of testing hours in our team, we determine not to use individual sl/tp. So we close by a global of legs. Being that we trade both ways to capture most movements. So. I do not know if a good ratio exists.
Regarding your question #3 in your pm.
Instead of increasing risk I have diminished it. The ring in question EU, GU, EG, I trade it 50/50/30 in mi live accounts. That is around 2-2.5% per month. And I can hold if fine if we have a similar negative correlation
EU-GU such like in Dec 2008/ feb 2009 and 2013 jan-April. LIke that I can work in peace in new angles of the concept without much monitor the accounts.
Yes, it is equity increase target. But equity target by leg individually.
Other are not really dangerous, but some rings could be very expensive to trade. Plus, each ring need a bit of statistical works to find historical ranges to determine a safe grid. After all, this is the key point of using price equations to build automated strategies. Calculated risk and very little deviation from backtests. Prices will always move in equation form.
I still have some Empty4 account in lives accounts. I still use one of the first original mql code. That is the easy part to implement.
Very interested to code this up as a JForex Strategy on Dukascopy and get it running on the Historical tester this weekend ...
Will post my results as they come to hand.
The two first (11-11-7, 20-20-7) with SL/TP for the two first pairs and a global target for the third pair (and all pairs of the combinations).
The third (50-50-30) without SL/TP, only a global target for the three pairs.
All configurations make crazy open drawdown! For the two first, the open drawdown is always at least equal to the closed profits... However for the last one, it is less important, so at a moment when you see that the open drawdown doesn't reduce, you manually close all the trades.
In conclusion : bigger boxes, no SL/TP.
I am working with EU-UCH-ECH and EG GU EU, buy-buy-sell, sell-sell-buy.
You do not have the required permissions to view the files attached to this post.
I can get +9% last month in JForex backtest on EUR/USD, USD/JPY, EUR/JPY tri-hedge
(2014-10-01 00:00:00 to 2014-11-01 00:00:00, Initial deposit 10000, Finish deposit 10933.75, Turnover 2287193.0, Comission 73.2) - but only 1% using same strategy with EUR/USD, USD/CHF, EUR/CHF. This is using a 10 pip basket profit target no SL.
BUT what I discovered was that I needed to base my BBS or SSB decision on whether the 'exotic' (EUR/JPY or EUR/CHF) was under or over its expected price based on the sum of the other two intruments.
My testing continues ...
fprophet » Mon Nov 24, 2014 4:52 am wrote:I can get +9% last month in JForex backtest on EUR/USD, USD/JPY, EUR/JPY tri-hedge
(2014-10-01 00:00:00 to 2014-11-01 00:00:00, Initial deposit 10000, Finish deposit 10933.75, Turnover 2287193.0, Comission 73.2) - but only 1% using same strategy with EUR/USD, USD/CHF, EUR/CHF. This is using a 10 pip basket profit target no SL.
BUT what I discovered was that I needed to base my BBS or SSB decision on whether the 'exotic' (EUR/JPY or EUR/CHF) was under or over its expected price based on the sum of the other two intruments.
My testing continues ...
Did you test that months when market moved 1000s of pips?
I am also trying jForex. Can tell more detail what box size you use for each pair and what direction?
Click here to go to the BoardKnowledgeBase