Phoenix is here! | Page 17 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Phoenix Forex Manager trading journal 25 replies

Phoenix Foundation 1 reply

Phoenix 2007 (new thread) 974 replies

Can someone fix the Phoenix EA? 0 replies

Here is a free Calculator...I don't know if its been posted here before 3 replies

Attachments: Phoenix is here!
Exit Attachments

Phoenix is here!

  • #321
  • Jul 16, 2006 3:00pm Jul 16, 2006 3:00pm
But it all depends on how much wins you have between now and then. If you have over 30 wins then you are already covered for those 3 losses.



Quoting zarni
Guys, sorry but you must be joking right...any system with an SL of 300 cannot be viable...THere will be an inevitable time when it hits 2 or 3 bad trades and that would be it...
  • #322
  • Jul 16, 2006 3:02pm Jul 16, 2006 3:02pm
Interesting thats a huge difference in results. Thanks for the info I will change to 30 to 300. Whats the lot that you used for your backtest? 1.00 or 0.1?



Quoting jhathawa
I backtested 25/250 on GBP from 1/1/06 to today and its basically break-even (you make $250 on 10k). 30/300 is up about 50% so there's quite a difference. I've found GBP and CHF to be the best candidates from backtests. I'm gonna demo 30/300 on GBP starting this week.
  • #323
  • Jul 16, 2006 3:05pm Jul 16, 2006 3:05pm
BenQ all the results I have seen so far shows closing trade total more than floating trade total.

Quoting BenQ
Hey Yardie, I agree, seems hard to ignore. But it also seems like we've done these settings before in the early days of firebird and it eventually crashed -- which is what gave rise to Phoenix. I guess the difference now would be no pipstep -- which really! hurt in the strong trends -- and the addition of the new SafePipsMinutes/Distance feature and UseCloseSignal. Hopefuly those along with the SafeArea will make a real difference this time. What concerns me though is his open orders can get quite huge. I'm not sure, but I think I noticed on one of his posted results that his negative open orders were almost as large as all the positive closed trades.
  • #324
  • Jul 16, 2006 3:20pm Jul 16, 2006 3:20pm
I don't know if you're interested in my opinion on this. I'm just gonna tell you anyways: ANY system with a SL of 300 is profitable in certain market situations. It works like this: well, we got our entry signal...hmmm...well, even if the signal (long or short) is wrong, it's not gonna hit the S/L because it's SOOOO far away. It's much more likely that it's gonna hit the T/P because it's 10 times nearer. However, the EntrySignal (generated by Phoenix's MA envelope) doesn't matter AT ALL for this kind of strategy. Why? Well, Phoenix's signal is short-term - it's a short-time breakout signal. That means it's doesn't say ANYTHING about the long-term trend - it doesn't even TRY to predict the long-term trend and it's really bad at this, too. So basically you could take ANY entrypoint (I have to confess that Phoenix's Entry points are a little better than random one, however for the long-term trend, they are just a tiny little bit better). With a 300 S/L & 30 T/P you would get nice profits even with a randomly chosen entry point (try some random-signal-EA's for proof), however, only in certain times! These times are the ranging periods. Because if there were huge trends - each more than 400 pips it would be just a 50-50 game if you get onto the right trend to not get stopped out but you don't get 300 pips, so a 50 win 50 loss game DOESN't work out with 30/300. However, in ranging markets, it's going up and down and eventually it's gonna hit your T/P. BUT, you need 10 wins on loss, that takes a while. So probably there's gonna be a period with big trends. And you lose BIG TIME!
Well, you don't have to listen to me but please think of it! And maybe you're lucky and it works so long till you have enough money to get out of it but I would NOT try it
  • #325
  • Jul 16, 2006 3:53pm Jul 16, 2006 3:53pm
Quoting Yardie
Interesting thats a huge difference in results. Thanks for the info I will change to 30 to 300. Whats the lot that you used for your backtest? 1.00 or 0.1?
Yardi,
I used 1.0 lots.

Zarni,
Yes, its a huge stop loss but I have downloaded the Alpari data to backtest on MT4 and no matter what time period I choose I get at least break even results (in EUR and JPY) and very good results with GBP and CHF. If you look back at my earlier post on results in this thread you'll see that various periods yield around 50% return on GBP and CHF. I agree it needs forward testing but if you stick to it for at least 6 months I think its viable. You'll notice the winners are far more than 10-1. closer to 20-1. As it stands if you use 18/42 or 25/50 this EA is not going to work if you trust backtesting at all. Its almost always break even. Maybe the next version will work with these TP/SL numbers......but for now I'm gonna demo with 30/300 and just see what happens.
  • #326
  • Jul 16, 2006 3:59pm Jul 16, 2006 3:59pm
Using Alpari Data I get really good results for 18/42, too. However, don't even think of taking backtesting for accurate! The history you're downloading is without ticks (at least I thin so) so the backtesting is not accurate AT ALL! Not even close.

Alright. 1 minutes reamaining. I wonder how Phoenix is doing this week. Last Friday was O.K => I made 4*18 pips, however nobody knows what's happening with the floats, right? Because I am down 25 pips on floats..
  • #327
  • Jul 16, 2006 4:09pm Jul 16, 2006 4:09pm
Quoting tradeigel
Using Alpari Data I get really good results for 18/42, too. However, don't even think of taking backtesting for accurate! The history you're downloading is without ticks (at least I thin so) so the backtesting is not accurate AT ALL! Not even close.

Alright. 1 minutes reamaining. I wonder how Phoenix is doing this week. Last Friday was O.K => I made 4*18 pips, however nobody knows what's happening with the floats, right? Because I am down 25 pips on floats..
I may be doing something wrong but with the standard settings (and Alpari data with every tick calculated) discussed on here the 18/42 EURO backtest results are not good. See attached. Since 1/1/06 to today I get a PF of 1.03 ($480 profit on 10k)....where did I differ from you on settings ?
Attached File(s)
File Type: zip StrategyTester_eur_15m_010106_071306.zip   10 KB | 294 downloads
  • #328
  • Jul 16, 2006 4:38pm Jul 16, 2006 4:38pm
Quoting jhathawa
I may be doing something wrong but with the standard settings (and Alpari data with every tick calculated) discussed on here the 18/42 EURO backtest results are not good. See attached. Since 1/1/06 to today I get a PF of 1.03 ($480 profit on 10k)....where did I differ from you on settings ?
Yeah, you know, nobody know yet if Phoenix is working on the long run...I haven't tried to backtest it on the whole year, just the last few weeks - The whole year take around 10 minutes here, how long does it take you? The results for the short time were those (as posted before) :
http://www.forexfactory.com/forexfor...&postcount=225

EDIT: Yeah, it really sucks on one year - I just backtested the first few months. I hope it doesn't perform like this in the future . But as said, backtesting stinks so much. So don't rely on it!
  • #329
  • Jul 16, 2006 4:41pm Jul 16, 2006 4:41pm
If so my mistake. I just don't like the huge sl. Are you planning on using all Mono's settings? Wasn't he using a 30 min chart with a wider envelope at .1 rather than .05? And if so, wasn't there an early consensus in the firebird thread that certain aspects of it were hard coded for a 15 min chart. I wonder if those 15 min attributes were carried over to Phoneix.

Quoting Yardie
BenQ all the results I have seen so far shows closing trade total more than floating trade total.
  • #330
  • Jul 16, 2006 4:46pm Jul 16, 2006 4:46pm
Quoting BenQ
If so my mistake. I just don't like the huge sl. Are you planning on using all Mono's settings? Wasn't he using a 30 min chart with a wider envelope at .1 rather than .05? And if so, wasn't there an early consensus in the firebird thread that certain aspects of it were hard coded for a 15 min chart. I wonder if those 15 min attributes were carried over to Phoneix.
No it's not hardcoded I think!
PHP Code
MA = iMA(NULL,MA_Timeframe,MA_Length,0,MODE_SMA,PRICE_OPEN,0);
The default for MA_Timeframe is 15 but you can change it!

PS: This PHP-tag sucks for MQL, we would need one for MQL4
  • #331
  • Jul 16, 2006 4:58pm Jul 16, 2006 4:58pm
Quoting tradeigel
Yeah, you know, nobody know yet if Phoenix is working on the long run...I haven't tried to backtest it on the whole year, just the last few weeks - The whole year take around 10 minutes here, how long does it take you? The results for the short time were those (as posted before) :
http://www.forexfactory.com/forexfor...&postcount=225

EDIT: Yeah, it really sucks on one year - I just backtested the first few months. I hope it doesn't perform like this in the future . But as said, backtesting stinks so much. So don't rely on it!
It takes 1-2 minutes for me I think to backtest 1/1/06 to today using every tick. My machine is pretty new though, 2.75GHZ.
  • #332
  • Jul 16, 2006 5:06pm Jul 16, 2006 5:06pm
What the ****? Is so fast...well, I got a 3200+, too. So it's not that bad either. It's probably because I use an emulator to run Windows Applications on Linux. But it shouldn't be sooo slow. I will look into this.
  • #333
  • Jul 16, 2006 5:16pm Jul 16, 2006 5:16pm
I don't know. In the early days of firebird there was talk of something being 15 Min.

And then maybe I thought of it again when I saw this quote from Hendrick a few posts ago.

"The indicator is hard-coded for TF=15M!"<?xml:namespace prefix = o ns = "urn:schemas-microsoft-com:office:office" /><o></o>



Quoting tradeigel
No it's not hardcoded I think!
PHP Code
MA = iMA(NULL,MA_Timeframe,MA_Length,0,MODE_SMA,PRICE_OPEN,0);
The default for MA_Timeframe is 15 but you can change it!

PS: This PHP-tag sucks for MQL, we would need one for MQL4
  • #334
  • Jul 16, 2006 5:27pm Jul 16, 2006 5:27pm
Well, I don't see anything hardcoded in Phoenix.
If you can change it with a variable it's not hard coded
  • #335
  • Jul 16, 2006 5:38pm Jul 16, 2006 5:38pm
Quoting jhathawa
Yardi,
I used 1.0 lots.

Zarni,
Yes, its a huge stop loss but I have downloaded the Alpari data to backtest on MT4 and no matter what time period I choose I get at least break even results (in EUR and JPY) and very good results with GBP and CHF. If you look back at my earlier post on results in this thread you'll see that various periods yield around 50% return on GBP and CHF. I agree it needs forward testing but if you stick to it for at least 6 months I think its viable. You'll notice the winners are far more than 10-1. closer to 20-1. As it stands if you use 18/42 or 25/50 this EA is not going to work if you trust backtesting at all. Its almost always break even. Maybe the next version will work with these TP/SL numbers......but for now I'm gonna demo with 30/300 and just see what happens.

Well I agree with you there may be periods it works..However the periods it doesn't work (and I believe those will come surely), it will blow up. If you notice recently the market does not know where to go. Once we get the trends from before it will be over.
The real way to test this EA is set your TP and SL to the same exact setting and see how it works over longer term. You cannot expect to make a sample test over limited data and conclude that it works based on limiting the probability of SL being hit by setting it to such a number and hoping market will turn back....
  • #336
  • Jul 16, 2006 5:41pm Jul 16, 2006 5:41pm
Quoting Yardie
But it all depends on how much wins you have between now and then. If you have over 30 wins then you are already covered for those 3 losses.
Right, and what happens if those 3 hit you on your real account when you turn the EA on on live trading?
  • #337
  • Jul 16, 2006 6:30pm Jul 16, 2006 6:30pm
Right, I understand you can change the variables. But if you've read the firebird thread -- and I know it's long -- I remember their was a lot of talk back about a month or so that something about firebird was hardcoded for M15. Maybe it was in the firebird doc from firebird's originator. No big, Hendrick re-wrote it anyway. It might have been in one of the early 63 versions. I don't know if you were trading it back then, but most of us were using M15 for a reason.

Quoting tradeigel
Well, I don't see anything hardcoded in Phoenix.

If you can change it with a variable it's not hard coded
  • #338
  • Jul 16, 2006 6:56pm Jul 16, 2006 6:56pm
I am testing Mono's strategy on the 15 minute time frame to see how it goes.

I agree with the earlier poster that say the default setting will lead to break even (or close to it).


Quoting BenQ
Right, I understand you can change the variables. But if you've read the firebird thread -- and I know it's long -- I remember their was a lot of talk back about a month or so that something about firebird was hardcoded for M15. Maybe it was in the firebird doc from firebird's originator. No big, Hendrick re-wrote it anyway. It might have been in one of the early 63 versions. I don't know if you were trading it back then, but most of us were using M15 for a reason.
  • #339
  • Jul 17, 2006 12:27am Jul 17, 2006 12:27am
Quoting zarni
Right, and what happens if those 3 hit you on your real account when you turn the EA on on live trading?
Then you have problems. I have also been running this EA on the 30min with at 30/300 but so far the market has not been trending in the longer term so it's hard to know how it will perform. So far since I began testing just over a month ago I have had all winners. The floating loss is scary at times and it does carry a huge risk. The only way to know how it will work in the long run is to keep demo testing it. I am not willing to risk real money until I see how it performs in a trending market.

Then again if you reduce your lot size so that the 300 pip risk is only 2 or 3% then 3 straight losses is no big deal. The profits will be smaller though but at least we'll survive longer in a "bad" market.

Wiz.
  • #340
  • Jul 17, 2006 12:42am Jul 17, 2006 12:42am
Quoting zarni
Guys, sorry but you must be joking right...any system with an SL of 300 cannot be viable...THere will be an inevitable time when it hits 2 or 3 bad trades and that would be it...
Put it this way. Most people would like to have it the other way around. High reward low risk. The problem is if you were aiming for 300 pip gains how often would you achieve them. Could you get 3 in a row with a risk of 30 pips? The logic says, low risk high reward but most of the time we lose. Having huge stops and small gains seems stupid but maybe, just maybe, that's what will make it work.

I've run quite a few tests with firebird. All of the 15 min ones bombed out and I thought the 30min with 30/300 was a stupid idea but it's the only one showing potential at the moment.

Please feel free to argue with me. I won't take offence. Maybe the way ahead is to do what most consider to be stupid.

My oppinion.

Wiz.
1 trader viewing now