Phoenix 2007 (new thread) - page 26
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Live trading with Phoenix
From the results Live trading below, 1 trade in profit was closed manualy otherwise would of end up losing and the last 3 lossing trades were closed also manualy otherwise would of lost the whole 84 pips.
It is too soon to go live with this EA. back testing and forward testing has also proven to be unsubstantiated in comparison to live trading results. I hope this would help you to develop this EA in a more controllable maner. I have heard some good idea's so far. I will wait and lookforward to see the up coming new versions.
Good luck to you all.
Ross.
forward testing
the results for this week's testing updated.
John
Hendrick,
I started testing this EA yesterday and I have a few questions. Perhaps this has already been convered in this thread, but why is there such a large disparity between my trades and the trades in your contest account? The datafeeds may be slightly different, but that still shouldn't account for such a big difference. Obviously, my initial trade is not going to be the same as yours since your position is open and mine would open later at a different time. But as time goes on and the EA generates more trades, will this disparity begin to disappear?
Also, have you used this EA on a live account, and if so, who do you find to be the best broker out there? I'm testing IBFX and FXDD right now, but I was wondering if there was a large difference between their live and demo feeds. I would appreciate any help.
Forward Test
Mine were facing 2 drawdown from different currency yesterday (AUDUSD, EURUSD), I will stop trading those 2 currencies, and continue with others.
The total profit is still positive, covered by USDCAD. The trade has been floating for 6 days (4 working days). So trading more than 2 days is not always ended with losses.
Within a week of test $1000 is becoming $1313 (30%)
Hendrick,
I started testing this EA yesterday and I have a few questions. Perhaps this has already been convered in this thread, but why is there such a large disparity between my trades and the trades in your contest account? The datafeeds may be slightly different, but that still shouldn't account for such a big difference. Obviously, my initial trade is not going to be the same as yours since your position is open and mine would open later at a different time. But as time goes on and the EA generates more trades, will this disparity begin to disappear?
Also, have you used this EA on a live account, and if so, who do you find to be the best broker out there? I'm testing IBFX and FXDD right now, but I was wondering if there was a large difference between their live and demo feeds. I would appreciate any help.For the contest, Hendrick were using this version. There is some limitation with the Lot amount, it can't be more than 5 lots at a time. So there is a limit in this EA.
And the parameter for USD/JPY is a bit different from Phoenix v4_2_03. I believe it is the latest updated parameters based on latest market situation.
I am not running your excellent EA yet Hendrik but I had some ideas regarding trade windows.
Due to the highly fractal nature of price movements, it may make sense to autoscale the time of the trade window, exit/stop signals, etc. A policy may also be written about days to hold a position with swap as a basis. Holding some positions will really hurt, due to swap costs. Why pay more to hold on a position that isn't moving?
Here is a Murrey Math student with a method for calculating interval to apply a fractal analysis method:
http://web.archive.org/web/20030115190850/http://www.webspace4me.net/~blhill/pages.aux/murrey/TKnotes.1.html
Pulled from the archives, it is step 3 and varies based on the price variation. There is also information that may be used to determine fractally scaled exit and stop positions.
The math seems simple enough, the harder part about a MM system appears to be zones of defense (2-1-2). Currently working with a physics friend on wavelet analysis and prediction to be combined with MM and Elliot waves. The math gets nasty fast, but the fractal scaling seems quite simple.
MM analysis without rules appears to be a price guessing system over time at this point in my studies - perfect for scaling Phoenix.
For the contest, Hendrick were using this version. There is some limitation with the Lot amount, it can't be more than 5 lots at a time. So there is a limit in this EA. And the parameter for USD/JPY is a bit different from Phoenix v4_2_03. I believe it is the latest updated parameters based on latest market situation.
Phoenix_4_CONTEST is the EA I have been testing.
I had to consider a reply for Phoenix Boosted quite carefully. The idea of trading corresponding currencies is somewhat appealing, the idea of making money 3 times instead of making money once has a basic fundamental appeal. I basically agree with all of the statements regarding money management, entry and exit signals being a mirror image or simply not useful.
I considered a few logic problems in order to determine how useful corresponding currencies would be in practice. The premise was a 100% exact currency, a 95% corresponding currency, and a time delayed version of each. (The mirrored numbers are essentially identical since there is no cost difference executing short or long in forex.)
With a 100% exact currency traded identically, you are basically taking 2 lots. There are a few differences: relative interest rates may be lower on one pair and increase or decrease swap costs, the pip spread on one currency pair could be lower than another, the dealer desk may not skew your results on minor pairs. (Time delays are treated separately)
Results for 100% exact:
If a trending move is made, and there are several 100% equivalent currencies to select, by all means choose the currency pair with the least swap cost.
If a quick exist position is taken, choose the currency pair with the least pip spread. (broker specific in some situations)
The second trading rule with a 1 pip spread would save 2 pips round trip. Average 10 trades per month where the strategy could be used and you just earned another 20 pips per month by executing the same trade differently.
At 95% correlation, each market needs to be evaluated separately for price, indicators, in addition to the calculations done above. If there is no time delay, then this is the real world that you trade in, with additional criteria that the 100% ideal doesn't account for. The exit, stop, trailing stop, etc all need to be adjusted. If there is a compelling reason on swap or pip spread, I would probably do this - but that data is all broker specific.
Lastly, time delays. If a news events hits one market at 9:30am and there is a movement at 9:30am one direction, trading in a corresponding currency may have lower volume, lower slippage, and similar movement. Just a thought, it would take a lot of research to verify this.
I had to consider a reply for Phoenix Boosted quite carefully. The idea of trading corresponding currencies is somewhat appealing, the idea of making money 3 times instead of making money once has a basic fundamental appeal. I basically agree with all of the statements regarding money management, entry and exit signals being a mirror image or simply not useful.
I considered a few logic problems in order to determine how useful corresponding currencies would be in practice. The premise was a 100% exact currency, a 95% corresponding currency, and a time delayed version of each. (The mirrored numbers are essentially identical since there is no cost difference executing short or long in forex.)
With a 100% exact currency traded identically, you are basically taking 2 lots. There are a few differences: relative interest rates may be lower on one pair and increase or decrease swap costs, the pip spread on one currency pair could be lower than another, the dealer desk may not skew your results on minor pairs. (Time delays are treated separately)
Results for 100% exact:
If a trending move is made, and there are several 100% equivalent currencies to select, by all means choose the currency pair with the least swap cost.
If a quick exist position is taken, choose the currency pair with the least pip spread. (broker specific in some situations)
The second trading rule with a 1 pip spread would save 2 pips round trip. Average 10 trades per month where the strategy could be used and you just earned another 20 pips per month by executing the same trade differently.
At 95% correlation, each market needs to be evaluated separately for price, indicators, in addition to the calculations done above. If there is no time delay, then this is the real world that you trade in, with additional criteria that the 100% ideal doesn't account for. The exit, stop, trailing stop, etc all need to be adjusted. If there is a compelling reason on swap or pip spread, I would probably do this - but that data is all broker specific.
Lastly, time delays. If a news events hits one market at 9:30am and there is a movement at 9:30am one direction, trading in a corresponding currency may have lower volume, lower slippage, and similar movement. Just a thought, it would take a lot of research to verify this.Hi Daraknor!
Thanks a lot for yor contribution. I've added your "article" to my list of "things to be investigate" (it's quite a list now!).
Thank you Hendrick!
I would also like to quip that hedge positions using correlating currencies only benefit the broker. If it really is a looser, don't wait to see how bad you can loose on it.