Invictus EAs | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

System Test: Invictus V Fratelli 541 replies

Trading the Invictus method 4 replies

One MT4, 2 EAs 2 replies

NewsTracker to control EAs around newstime 4 replies

EAs and losing connection 3 replies

Attachments: Invictus EAs
Exit Attachments

Invictus EAs

  • #1
  • First Post: Nov 16, 2011 11:33pm Nov 16, 2011 11:33pm
here is the thread of the INVICTUS EA's posted in the INVICTUS thread: http://www.forexfactory.com/showthread.php?p=5115718

The EA's are posted here:

http://www.forexfactory.com/showthre...25018&page=105

post 1566

and here:

http://www.forexfactory.com/showthre...25018&page=105

post 1568


As I would like to keep the thread clear from any EA discussion, I have opened this one, for the ones interested in the EA development.

Green pips to all

Udine
Sleep less, live fast, as it aint gonna last
  • #2
  • Nov 17, 2011 1:00am Nov 17, 2011 1:00am
Hi Udine... I am hoping this EA Thread will ease us to get more green pips. :-)
  • #3
  • Nov 17, 2011 1:43am Nov 17, 2011 1:43am
Quoting fxolive
Hi Udine... I am hoping this EA Thread will ease us to get more green pips. :-)

fxolive,

I have done nothing, I am not a coder, and doing fine trading manually

I just opened this thread and posted the links to the EA's other people have coded, as I want to keep the INVICTUS thread focussed on manual trading.

Udine
Sleep less, live fast, as it aint gonna last
  • #4
  • Nov 17, 2011 2:07am Nov 17, 2011 2:07am
Quoting Udine
fxolive,

I have done nothing, I am not a coder, and doing fine trading manually

I just opened this thread and posted the links to the EA's other people have coded, as I want to keep the INVICTUS thread focussed on manual trading.

Udine
Thanks for opening this EA thread, Udine. It is greatly appreciated that you routed the EA discussion out of the main forum for INVICTUS.

I anticipate that not more than 5 people will get a profitable result; more realistically, I expect to see only 1. But as a scientist, I keep an open mind and draw conclusions only after sufficient facts are available.

DrDave
To free Gazans of Hamas, use whatever it takes.
  • #5
  • Nov 17, 2011 7:13am Nov 17, 2011 7:13am
Quoting Udine
here is the thread of the INVICTUS EA's posted in the INVICTUS thread: http://www.forexfactory.com/showthread.php?p=5115718

The EA's are posted here:

http://www.forexfactory.com/showthre...25018&page=105

post 1566
This one need smInvictus AtrRangeFilter_v3.ex3 Indicator.
  • #6
  • Nov 17, 2011 7:21am Nov 17, 2011 7:21am
Quoting sissi
This one need smInvictus AtrRangeFilter_v3.ex3 Indicator.
Why?

it has not take any trades yet
  • #7
  • Nov 17, 2011 7:24am Nov 17, 2011 7:24am
Quoting joancb
Why?
http://easycaptures.com/fs/uploaded/621/8141812738.png
  • #8
  • Nov 17, 2011 5:42pm Nov 17, 2011 5:42pm
Hi All,

Posting the last version of the EA and indicators:
Utctime.mqh goes to experts/include folder and it is needed for the EA to work.
Inv_EA.mq4 is the EA source code
The rest of it are indicators. Out of them, the EA only needs FFCal.mq4. This is a news indicator I got from http://www.forexfactory.com/showthre...=19293&page=40
Inv_has_quick.mq4 is the non-repainting MTF HA indicator which displays arrows.
All others are indicators which display pivot and round number levels.

Changes: many small fixes, added DayBeginsAtUTC parameter which sets a UTC hours when a new day starts. If you set it to 22, for instance, this will mean the new day starts at NY close (provided you set BrokerOffsetFromUTC correcty). Also added HaColorFilter, which means if we are waiting for a breakout after an arrow and current candle HA color changes against us, no trade. Looks like Udine is using this rule.

The EA currently trades exactly to the rules. The difference between its peformance for different brokers has been minimized. The only source of difference are H4 candles which may be formed differently depending on the broker.
I must say that I'm not impressed with the performance of this strategy. It shows some promise, especially on EU, but I wouldn't trust it my money.
As I said previously, if one wants to trade it, I would recommend recoding it for Dukascopy JForex or maybe Multicharts and giving it a good test on proper tick data. MT4 backtesting cannot be trusted.


Thanks,
Mcs
Attached File(s)
File Type: mqh utctime.mqh   3 KB | 361 downloads
File Type: mq4 Inv_EA.mq4   29 KB | 388 downloads
File Type: mq4 FFCal.mq4   50 KB | 360 downloads
File Type: mq4 inv_has_quick.mq4   24 KB | 446 downloads
File Type: mq4 inv_other_pivots.mq4   4 KB | 352 downloads
File Type: mq4 inv_resistance_pivots.mq4   4 KB | 349 downloads
File Type: mq4 inv_support_pivots.mq4   4 KB | 338 downloads
  • #9
  • Nov 17, 2011 11:16pm Nov 17, 2011 11:16pm
Quoting mcs
Hi All,

Posting the last version of the EA and indicators:
Utctime.mqh goes to experts/include folder and it is needed for the EA to work.
Inv_EA.mq4 is the EA source code
The rest of it are indicators. Out of them, the EA only needs FFCal.mq4. This is a news indicator I got from http://www.forexfactory.com/showthre...=19293&page=40
Inv_has_quick.mq4 is the non-repainting MTF HA indicator which displays arrows.
All others are indicators which display pivot and round number levels.

Changes: many small fixes, added DayBeginsAtUTC...
Good morning MCS,

GREAT JOB, I'll try it out !!

Again well done.

Udine
Sleep less, live fast, as it aint gonna last
  • #10
  • Nov 17, 2011 11:48pm Nov 17, 2011 11:48pm
Quoting mcs
Hi All,

Posting the last version of the EA and indicators:
Utctime.mqh goes to experts/include folder and it is needed for the EA to work.
Inv_EA.mq4 is the EA source code
The rest of it are indicators. Out of them, the EA only needs FFCal.mq4. This is a news indicator I got from http://www.forexfactory.com/showthre...=19293&page=40
Inv_has_quick.mq4 is the non-repainting MTF HA indicator which displays arrows.
All others are indicators which display pivot and round number levels.

Changes: many small fixes, added DayBeginsAtUTC...
Thanks a lot i will try

I want to use the inv_has_quick.mq4 in m5 but no bars will show.
Can you fix is> or is this a problem in my MT4
  • #11
  • Nov 18, 2011 12:08am Nov 18, 2011 12:08am
Quoting sissi
Thanks a lot i will try

I want to use the inv_has_quick.mq4 in m5 but no bars will show.
Can you fix is> or is this a problem in my MT4
It works for me on M5. Maybe you don't have enough bars on the chart? The number of bars should be > 2 * MaxBars, which is 2000 by default, I think.
  • #12
  • Nov 18, 2011 12:09am Nov 18, 2011 12:09am
Quoting mcs
It works for me on M5. Maybe you don't have enough bars on the chart? The number of bars should be > 2 * MaxBars, which is 2000 by default, I think.
Ah thanks. Now it works for me
  • #13
  • Nov 20, 2011 8:44pm Nov 20, 2011 8:44pm
Hi All,

Posting the last versions of the EA and the indicator. The indicator had a bug resulting in (rarely) incorrect historical data arrows. Now fixed. Some further additions to the EA include closing the position after getting a different color HA candle and break even level.

If anybody wants to use the indicator in their code, it is done like this:

Alert(iCustom(NULL, 0, "inv_has_quick", /*indicator parameters*/"", "", "", 1/*broker offset from UTC*/, "", 7, 0, 17, 0, 0, 0, 1, 1, "", 0 /*day begins at UTC*/,
"", true/*room/range filter*/, 0.2, "", true/*ATR*/, 5, 25, "", true/*pivot*/, "", 10.0, 0.0, "", true/*roundnumber filter*/, 10.0, 0.0,
"", true/*last bar color filter*/, "", true/*news filter*/, 30, 30, "", false, true, true, 1000/*MaxBars*/, 20/*last parameter - arrow spacing*/,
3/*buffer index*/, 1/*bar index*/));

Make sure you have at least MaxBars * 2 bars loaded first.

Buffer indexes description:
0 - Period1: up signal if > 0, no signal if == 0
1 - Period1: down signal if > 0, no signal if == 0
2 - Period2 up
3 - Period2 down
4 - Period3 up
5 - Period3 down


I think I'm done with this system for now. My verdict: it may be good for manual trading with experienced traders (who have sixth sense of the market), but as just a mechanical system it is barely profitable. Simple optimizations around the rules of the system do not help, in order to become profitable it should be modified seriously to the degree when it is a totally new system.

Thanks,
Mcs
Attached File(s)
File Type: mq4 inv_has_quick.mq4   23 KB | 298 downloads
File Type: mq4 Inv_EA.mq4   29 KB | 293 downloads
  • #14
  • Nov 20, 2011 8:59pm Nov 20, 2011 8:59pm
Quoting mcs
Hi All,

I think I'm done with this system for now. My verdict: it may be good for manual trading with experienced traders (who have sixth sense of the market), but as just a mechanical system it is barely profitable. Simple optimizations around the rules of the system do not help, in order to become profitable it should be modified seriously to the degree when it is a totally new system.

Thanks,
Mcs

Thanks MCS. Great effort. But it is shame that you can not make this system profitable.

One question? Does your new HAS quick indicator repaint? For some reason, I saw once in one-minute chart, the 5-minute histogram repainted one bar back. Maybe just my mistake. I will check it further.

Anyway. Thanks for the effort to confirm my believe that this system as is would not be profitable mechanically.
  • #15
  • Nov 20, 2011 9:19pm Nov 20, 2011 9:19pm
Hi precisiontra,

HAS indicator doesn't repaint (it was the whole point of its creation), there was a bug though which could result in repainting under rare circumstances, which is fixed now.
Has it been a repainting one, it would be very easy to get great results for backtesting

I did a proper backtesting using tick history in JForex, and within the original rules I couldn't get PF more than 1.3. There is still some room for experiments, but so far the results are not encouraging.
Of course, I can change the rules quite a bit and probably boost the PF, but it would be almost completely new system then.


Cheers,
Mcs

Quoting precisiontra
Thanks MCS. Great effort. But it is shame that you can not make this system profitable.

One question? Does your new HAS quick indicator repaint? For some reason, I saw once in one-minute chart, the 5-minute histogram repainted one bar back. Maybe just my mistake. I will check it further.

Anyway. Thanks for the effort to confirm my believe that this system as is would not be profitable mechanically.
  • #16
  • Nov 20, 2011 9:38pm Nov 20, 2011 9:38pm
Quoting mcs
Hi precisiontra,

HAS indicator doesn't repaint (it was the whole point of its creation), there was a bug though which could result in repainting under rare circumstances, which is fixed now.
Has it been a repainting one, it would be very easy to get great results for backtesting
Cheers,
Mcs
Thanks MCS for the prompt reply.

Yep, maybe that small bug in the previous version you mentioned could result in repainting under rare circumstances. I believe that's what I saw last Friday in a one-minute chart. Thanks for fixing this bug. I will try live the newer version.

I guess you would not mind that I change your code/algorithm a bit to suit my needs right?. I've studying your code the other day, and I think there are some minor discrepancies of your 4-hour HAS histogram bars compared to my non-repainting HAS histogram bars. I used a completely different algorithm.

Although the differences are minor, but I believe it could change the back-test results. Again, it could be my mistake. Now, since you've fixed the re-painting issue, I could live compare the two HAS and see why that discrepancies occur once in a while during the choppy/range markets. In a trending phase, there are no problems.

Again, thanks for fixing the bug. I will need to change your codse a bit to have a fair comparison.
  • #17
  • Nov 20, 2011 9:45pm Nov 20, 2011 9:45pm
Of course you are welcome to change the code and use it as you want. This is exactly the reason why I posted it, so people are able to have a look and play with it. Maybe someone even gets something profitable out of it, who knows?
That H4 discrepancies were almost surely related to that bug, I also noticed them when comparing two versions of the indicator. They should be gone now. Please let me know should you notice any other discrepancies.


Thanks,
Mcs


Quoting precisiontra
Thanks MCS for the prompt reply.

Yep, maybe that small bug in the previous version you mentioned could result in repainting under rare circumstances. I believe that's what I saw last Friday in a one-minute chart. Thanks for fixing this bug. I will try live the newer version.

I guess you would not mind that I change your code/algorithm a bit to suit my needs right?. I've studying your code the other day, and I think there are some minor discrepancies of your 4-hour HAS histogram bars compared to my non-repainting HAS histogram bars. I used...
  • #18
  • Nov 20, 2011 10:05pm Nov 20, 2011 10:05pm
Quoting mcs
Of course you are welcome to change the code and use
That H4 discrepancies were almost surely related to that bug, I also noticed them when comparing two versions of the indicator. They should be gone now. Please let me know should you notice any other discrepancies.


Thanks,
Mcs


Great. Now I feel much better now. For a while I thought that my version of the MTF HAS non-repainting histogram had problems. Great, and Thanks.

I've used my own version of MTF HAS for almost a year now, with some other non-repainting momentum MTF indicators for trend and momentum confirmation, and I was getting pretty good results when live manual trading.

Of course, since this is a trend/momentum driven system, finding a trending market is more important that all the HAS bars have to align with the same color. I firmly believe in your HAS EA, if you add a non-repainting MTF momentum indicator, such as MTF ADX with BollingerBand expansion momemtum indicator or MTF Explosion indicator (from MT4 code base) for trend/momentum confirmation, you may be able to turn your HAS EA into profitability.

Of course these momentum indicators that I mentioned in the present MTF form will re-Paint when displayed in real-time on a lower time-frame just like all other MTF indicators.

I also use the Better Volume indicator for confirmation. I also re-wrote the the Better Volume indicator to be MTF and Multi-Symbol. So, you can display the Better Volume histogram of 1 Hour AUDUSD and EURUSD on a 15 minute EURUSD chart, which will give you a great confirmation of the ATR/Explosion break-outs. Just some of the things I consider when I trade my HAS system. I do not care about the Pivots, Round Number, Lunch time break, Coffer times, etc. When a real breakout occurs with MTF Volume and MTF Momentum confirmation, it will break all pivots, round numbers, S/Rs and forget about lunch time or coffee break ;-). The trend will just continue and never stop.

Just some more thoughts if you ever want to improve your next HAS EA. If not, I may continue your EA work with my system and see what happens. At least you have a good start. Thanks.

PrecisionTrader.
  • #19
  • Nov 20, 2011 11:44pm Nov 20, 2011 11:44pm
Thank you very much for your suggestions, PrecisionTrader. I will have a look at improving this system later.
You are right, those pivot levels and coffee breaks do not work (everybody calculates them differently anyway). Round numbers are a bit better because of 'option barrier' thing, but not by much.

Cheers,
Mcs

Quoting precisiontra
Great. Now I feel much better now. For a while I thought that my version of the MTF HAS non-repainting histogram had problems. Great, and Thanks.

I've used my own version of MTF HAS for almost a year now, with some other non-repainting momentum MTF indicators for trend and momentum confirmation, and I was getting pretty good results when live manual trading.

Of course, since this is a trend/momentum driven system, finding a trending market is more important that all the HAS bars have to align with the same color. I firmly believe in your HAS...
  • #20
  • Nov 21, 2011 1:44pm Nov 21, 2011 1:44pm
Previously posted EA had a bug when using HaClose. Fixed this.
Also added a different filter for breakouts - now if HaColorFilter == 2 the EA after the signal waits (maybe for several candles) till either a breakout happens or opposite HA color candle fully formed. With HaColorFilter == 1 the filtering works as previously - waits for breakout only during the next candle and cancel the signal if this candle changes HA color to opposite while it is formed.
HaColorFilter == 0 means that the EA waits for the breakout during the next candle immediately after signal and ignores its HaColor.
Attached File(s)
File Type: mq4 Inv_EA.mq4   29 KB | 333 downloads
0 traders viewing now