Phoenix 2007 (new thread) | Page 12 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Phoenix Forex Manager trading journal 25 replies

Phoenix Foundation 1 reply

Can someone fix the Phoenix EA? 0 replies

Automated Trading Championship 2007: New Challenge! 33 replies

Phoenix is here! 848 replies

Attachments: Phoenix 2007 (new thread)
Exit Attachments

Phoenix 2007 (new thread)

  • #221
  • Oct 16, 2006 7:36am Oct 16, 2006 7:36am
comrades,

i'm pretty new here and trying to learn the ins and outs of expert advisors

i seem to have managed to get the phoenix ea running using the metatrader 15 min historic data as suggested but when i run phoenix on the old data it tells me i run out of money

am i doing something wrong or does phoenix fail to profit using last two years of data?
  • #222
  • Oct 16, 2006 8:41am Oct 16, 2006 8:41am
Hi Hendrick,

<o></o>You mentioned that optimizing of <st1:City w:st="on"><st1lace w:st="on">Phoenix</st1lace></st1:City> is not easy job. It looks so with so many parameters.
<o></o>I am not a programmer guru at all. Also I am pretty busy during day.
But if there is something I can do, it is performing back tests on my home computer.<o></o>
Feel free to send us/me any range of settings you would like to test and we can help you with that peace of work.<o></o>

<o> </o>-snoopy<o></o>
  • #223
  • Oct 16, 2006 8:51am Oct 16, 2006 8:51am
Hmmm, I had a position closed out with the first release of PBooster on the GBPJPY for about $250. Would the new version of PBooster entered and exited this trade at the same rate? Just wondering if I should count this one in the forward testing.
  • #224
  • Oct 16, 2006 9:09am Oct 16, 2006 9:09am
Quoting Palmer
Hmmm, I had a position closed out with the first release of PBooster on the GBPJPY for about $250. Would the new version of PBooster entered and exited this trade at the same rate? Just wondering if I should count this one in the forward testing.
I had the same trade happen. Still trying to figure it out. I've still not loaded the newest since the last bug.
  • #225
  • Oct 16, 2006 10:49am Oct 16, 2006 10:49am
A short recap on <?xml:namespace prefix = st1 ns = "urn:schemas-microsoft-com:office:smarttags" /><st1lace w:st="on"><st1:City w:st="on">Phoenix</st1:City></st1lace> 4.2 and Phoenix Booster.<?xml:namespace prefix = o ns = "urn:schemas-microsoft-com:office:office" /><o></o>

<o> </o>

<st1:City w:st="on"><st1lace w:st="on">Phoenix</st1lace></st1:City> 4.2<o></o>

<o> </o>

After testing for several weeks I came to the conclusion that <st1:City w:st="on"><st1lace w:st="on">Phoenix</st1lace></st1:City> 4.2 is making more profitable than losing trades. So, on the long run it should be profitable after all. Only problem is that the SL of 84 (TP:SL is 1:2) has a too great impact on your account. First I thought it was impossible to find a better TP/SL ratio, but I’ve changed my mind. The good news is, that there are better TP/SL ratios. I’m now testing several ratios (give me some time, because there’s more than <st1:City w:st="on"><st1lace w:st="on">Phoenix</st1lace></st1:City> in my life: family, work etc.). I think with these ratios the performance of <st1:City w:st="on"><st1lace w:st="on">Phoenix</st1lace></st1:City> will improve substantially. <o></o>

<o> </o>

<st1:City w:st="on"><st1lace w:st="on">Phoenix</st1lace></st1:City> Booster<o></o>

<o> </o>

I have high hopes for Booster! It’s functioning now but only half. As soon as trades are placed for the base currency and its correlated currencies the other base and correlated currencies don’t trade. I know why and will fix this as soon as possible. I will make a new version were we can all test are own base and correlated currencies. <o></o>

<o> </o>

But now I need your help: if all the regular visitors of this forum will be so kind to search the internet for “currency correlation” and try to give groups of currencies that are strongly correlated to each other and post these in this forum. I really would appreciate this!<o></o>

<o> </o>

<st1:City w:st="on"><st1lace w:st="on">Phoenix</st1lace></st1:City> is on the right track. It only needs to be improved. If we work all together then I’m sure we will succeed at last! Come on, there’s work to do!<o></o>

<o> </o>
  • #226
  • Oct 16, 2006 10:55am Oct 16, 2006 10:55am
Hendrick,

wouldn't it make sense to add this to the code of Phoenix & P.Booster:

right in the beginning:
#define version <PUT HERE THE VERSION NUMBER (like v2.2)>

string EAComment="Phoenix "+version;

and in each Order*()-function put only EAComment as comment for the order. This way we can keep track of which version opened / closed positions. I like that so I always have to add it to each version
  • #227
  • Oct 16, 2006 11:06am Oct 16, 2006 11:06am
Quoting Hendrick
A short recap on <?xml:namespace prefix = st1 ns = "urn:schemas-microsoft-com:office:smarttags" /><st1lace w:st="on"><st1:City w:st="on">Phoenix</st1:City></st1lace> 4.2 and Phoenix Booster.<?xml:namespace prefix = o ns = "urn:schemas-microsoft-com:office:office" /><o></o>

<o></o>

<st1:City w:st="on"><st1lace w:st="on">Phoenix</st1lace></st1:City> 4.2<o></o>

<o></o>

After testing for several weeks I came to the conclusion that <st1:City w:st="on"><st1lace w:st="on">Phoenix</st1lace></st1:City> 4.2 is making more profitable than losing trades. So, on the long run it should be profitable after all. Only problem is that the SL of 84 (TP:SL is 1:2) has a too great impact on your account. First I thought it was impossible to find a better TP/SL ratio, but I’ve changed my mind. The good news is, that there are better TP/SL ratios. I’m now testing several ratios (give me some time, because there’s more than <st1:City w:st="on"><st1lace w:st="on">Phoenix</st1lace></st1:City> in my life: family, work etc.). I think with these ratios the performance of <st1:City w:st="on"><st1lace w:st="on">Phoenix</st1lace></st1:City> will improve substantially. <o></o>

<o></o>

<st1:City w:st="on"><st1lace w:st="on">Phoenix</st1lace></st1:City> Booster<o></o>

<o></o>

I have high hopes for Booster! It’s functioning now but only half. As soon as trades are placed for the base currency and its correlated currencies the other base and correlated currencies don’t trade. I know why and will fix this as soon as possible. I will make a new version were we can all test are own base and correlated currencies. <o></o>

<o></o>

But now I need your help: if all the regular visitors of this forum will be so kind to search the internet for “currency correlation” and try to give groups of currencies that are strongly correlated to each other and post these in this forum. I really would appreciate this!<o></o>

<o></o>

<st1:City w:st="on"><st1lace w:st="on">Phoenix</st1lace></st1:City> is on the right track. It only needs to be improved. If we work all together then I’m sure we will succeed at last! Come on, there’s work to do!<o></o>

<o></o>
Hi: Handrick

First thank you very much. and wish the best luck in life and winning this contest.
Just as start tack a look at the following site:

http://www.mataf.net/en/analysis-correlation.htm

Best Regards
Yankey
  • #228
  • Oct 16, 2006 11:16am Oct 16, 2006 11:16am
Quoting Yakovzvi
Hi: Handrick

First thank you very much. and wish the best luck in life and winning this contest.
Just as start tack a look at the following site:

http://www.mataf.net/en/analysis-correlation.htm

Best Regards
Yankey
Hi Yankey,

Thanks for the link!
  • #229
  • Oct 16, 2006 11:28am Oct 16, 2006 11:28am
Quoting Hendrick

But now I need your help: if all the regular visitors of this forum will be so kind to search the internet for “currency correlation” and try to give groups of currencies that are strongly correlated to each other and post these in this forum. I really would appreciate this!
<o></o>
Hi Hendrick,

at Oanda's Homepage there is also a good site about currency correlation.
Please try this one:

http://fxlabs.oanda.com/cgi/fxlabs.pl?id=4

Wish you the best
mamba22
  • #230
  • Oct 16, 2006 11:33am Oct 16, 2006 11:33am
Quoting saltcafe
I've been thinking about looking at ways of optimizing <?xml:namespace prefix = st1 /><st1lace w:st="on"><st1:city w:st="on">Phoenix</st1:city></st1lace> and have thought of something different.

Nearly all trades that <st1lace w:st="on"><st1:city w:st="on">Phoenix</st1:city></st1lace> takes always take a loss initially. That is to be expected and I don't have a problem with that at all. However what may be of value is to optimize the orders being placed to take advantage of this.

For example currently phoenix at the time a signal is generated, would take a buy order at 1.2600 and set TP @ 1.2640 and SL of 1.2520. Roughly a 1:2 ratio, and I've been thinking about reducing this ratio to roughly 1:1. (up to 80 pips loss for a possible 40 pips gain)

So a new proposal would be at 1.2600 (when the signal is generated) place a limit order to buy at 1.2580 with a TP of 1.2640 and SL 1.2520. This gives us a 1:1 ratio for the trade. (up to 60 pips loss for a 60 pips gain)

Now this gives us a couple of things:

a) we have reduced the level of loss, (80 pips down to 60 pips) therefore for the same order, without increasing the risk profile at all, we can actually take a larger position

b) also if the trades do play out as expected then we increase the profit factor (due to larger amount of pips gained and larger position size)

c) however, we increase the risk that a trade won't be taken at all, because our new order limit levels may not be reached so we may miss out on the profitable 40 pips

What do you think? Does this sound like a good avenue to pursue for this EA?<?xml:namespace prefix = o /><o></o>
Hi Saltcafe,

I really appreciate your contributions to this thread. Great stuff and very helpfull!! But I don't agree with this proposal. It will highly change the concept of Phoenix (Phoenix is a counter-trend EA and not a trend follower). I agree about your opnion to reduce the level of loss. But I think there are other solutions for that (read my previous post about changing the SL:TP ratios)
  • #231
  • Oct 16, 2006 11:34am Oct 16, 2006 11:34am
Quoting nvwine
Hendrick,
Once again I'm not sure if this is a problem. Is Booster only supposed to take a trade on one pair at a time?
Since Booster took the 3 trades shown in post #210 the EA no longer goes through the 'signals' and the 'buy and sell conditions' part of the code. Before it took the 3 positions I had the information shown in the zip file below when I looked at the experts tab. Now I get no information even with the new version. This makes me think that now that it opened a trade it is now skipping this whole area of the code. I added some testing info to your code along with a couple of bool operators at the top on a temporary basis because I like to see what is happening. You can see it below at the bottom of the text file in the zip file. I realize it may be ugly but it is the best I can do.

As I said above now I get nothing. Looks like this for 60 minutes.

Look in the zip file for how it used to update before booster opened 3 out of 4? trades.
Hi Nvwine,

Wait for Booster version 3!
  • #232
  • Oct 16, 2006 11:36am Oct 16, 2006 11:36am
Quoting snoopy
Hi Hendrick,

<?xml:namespace prefix = o /><o></o>You mentioned that optimizing of <?xml:namespace prefix = st1 /><st1:City w:st="on"><st1lace w:st="on">Phoenix</st1lace></st1:City> is not easy job. It looks so with so many parameters.
<o></o>I am not a programmer guru at all. Also I am pretty busy during day.
But if there is something I can do, it is performing back tests on my home computer.<o></o>
Feel free to send us/me any range of settings you would like to test and we can help you with that peace of work.<o></o>

<o></o>-snoopy<o></o>
Thanks Snoopy. I'm still searching for the perfect EURGBP settings!
  • #233
  • Oct 16, 2006 11:37am Oct 16, 2006 11:37am
Quoting mamba22
Hi Hendrick,

at Oanda's Homepage there is also a good site about currency correlation.
Please try this one:

http://fxlabs.oanda.com/cgi/fxlabs.pl?id=4

Wish you the best
mamba22
That's a good one!
  • #234
  • Oct 16, 2006 11:39am Oct 16, 2006 11:39am
Quoting Palmer
Hmmm, I had a position closed out with the first release of PBooster on the GBPJPY for about $250. Would the new version of PBooster entered and exited this trade at the same rate? Just wondering if I should count this one in the forward testing.
Hey Palmer!

Booster is still very buggy. Wait for version 3! Will be much, much better!
  • #235
  • Oct 16, 2006 11:49am Oct 16, 2006 11:49am
Will do...Is it me or is Phoenix not performing well wil FXDD??? I had several open positions before I left for work, most in the negative. I'll find out later. It just looks as if some other brokers are a lot more P4 friendly.
  • #236
  • Oct 16, 2006 2:12pm Oct 16, 2006 2:12pm
Hendrick,

Another link to more currency correlation chart info.

http://www.mataf.net/en/analysis-correlation.htm

Wackena
  • #237
  • Oct 16, 2006 2:43pm Oct 16, 2006 2:43pm
I know this is meant to be an EA, but is there anyone who would be willing to turn this into an indicator since my account is with someone who doesn't use MT4, but I have the access to my computer 24/6 so I can still get into trades rather easily? I can finish up writing in the piece of code to have it SMS you with which currency pair, entry, exit, s/l so you can try and get in whenever you can, but I need someone else with a bit more coding experience to get me started? If anyone wants to take this on, I would appreciate it. After it's all started, I wouldn't mind keeping the indicator version up to date with the EA version for those interested (I can modify code, I'm just not so good at creating it out of nothing).
  • #238
  • Oct 16, 2006 2:48pm Oct 16, 2006 2:48pm
Phoenix is kicking butt at the EA championships! #2 right now. Great job Hendrick!

http://championship.mql4.com/2006/participants
  • #239
  • Oct 16, 2006 2:49pm Oct 16, 2006 2:49pm
Well...to curreny correlation:
Maybe it's an idea to read out current currency correlation and then let Booster act accordingly.

to today's performance:
It's just AWESOME!!
Opened 8 trades, 4 opened by Booster, 4 by regular Phoenix. Only 1 (!) was closed in loss the other 7 (!) went into profit! That's pretty much like a moneyMakingMachine , only wish it was live acc. - will probably do that pretty soon.
Thanks for Phoenix, Hendrick.
  • #240
  • Edited 3:27pm Oct 16, 2006 2:54pm | Edited 3:27pm
Hendrick,
Another link. This has method below on how to calculate correlation.

http://www.investopedia.com/articles.../05/051905.asp

Here is the correlation-calculation process reviewed step by step:

  1. <LI nd="12">Get the pricing data for your two currency pairs; say they are GBP/USD and USD/JPY <LI nd="13">Make two individual columns, each labeled with one of these pairs. Then fill in the columns with the past daily prices that occurred for each pair over the time period you are analyzing <LI nd="14">At the bottom of the one of the columns, in an empty slot, type in =CORREL( <LI nd="15">Highlight all of the data in one of the pricing columns; you should get a range of cells in the formula box.
  2. Type in comma
  3. Repeat steps 3-5 for the other currency
  4. Close the formula so that it looks like =CORREL(A1:A50,B1:B50)
  5. The number that is produced represents the correlation between the two currency pairs

Most articles I have read states that currency correlations are dynamic and always changing in some degree over time. Can EA be linked, via a script to price history file? Then current correlations could be calculated and EA could adjust order trigger routines accordingly.

Wackena

1 trader viewing now