Forex Factory (https://www.forexfactory.com/forum.php)
-   Trading Systems (https://www.forexfactory.com/forumdisplay.php?f=71)
-   -   My "Ultron" EA for GBPUSD H1 timeframe (https://www.forexfactory.com/showthread.php?t=840339)

SigmaEnigma Mar 15, 2019 5:09pm | Post# 1161

{quote} By default, the authorís version of this system uses time 6 and 21 ó this is for GMT. In version LITE_3, I made default 8 and 23 - this is for GMT + 2. Accordingly, if you use GMT 0, then try applying time 6 and 21.
Because I was using PaulDV's EA, I was sticking to the 7 and 19 time-frame he set specifically for that EA which are the default values for GMT servers. I think those numbers are right, I'm just confused as to why others got a buy signal when on my end the EA was out of trading hours. So did it work as intended then? Or are 7 and 19 not the correct values I should be using? I'm using Forex.com as my broker.

cdsaa Mar 15, 2019 5:15pm | Post# 1162

{quote} Because I was using PaulDV's EA, I was sticking to the 7 and 19 time-frame he set specifically for that EA which are the default values for GMT servers. I think those numbers are right, I'm just confused as to why others got a buy signal when on my end the EA was out of trading hours. So did it work as intended then? Or are 7 and 19 not the correct values I should be using? I'm using Forex.com as my broker.
Looks like you are out of trading hours. We all got a trade.

Some of us are on broker with +3 !and some on +2

SigmaEnigma Mar 15, 2019 6:04pm | Post# 1163

I guess moving forward I'll be using 6 and 21 for the time settings for PDV_1031. It makes a few more trades but the profitability remains the same (testing from Jan 1st to today, 2019). Net profit is a bit higher from those extra trades anyway. Welp, good luck with the buy order for everyone who got it, I guess I'm sitting this one out lol

PaulDV Mar 15, 2019 6:59pm | Post# 1164

{quote} Because I was using PaulDV's EA, I was sticking to the 7 and 19 time-frame he set specifically for that EA which are the default values for GMT servers. I think those numbers are right, I'm just confused as to why others got a buy signal when on my end the EA was out of trading hours. So did it work as intended then? Or are 7 and 19 not the correct values I should be using? I'm using Forex.com as my broker.
I've just checked, and Forex.com appear to have a server time of GMT-12 (https://www.myfxbook.com/forex-broker/forex-com/10).

The EA checks the trading hours set against the current broker's time. So when our orders opened at 17:00 GMT, your broker's server time would have been 05:00, which would explain why it didn't open an order.

I would recommend double-checking your broker's offset and using the new Broker Server Offset parameter to obtain the correct time window. e.g. if your broker offset really is -12, then set the start and end hours to the default of 07 and 19, and then enter -12 for the broker offset and you should be good to go.

EDIT: I've made a small change to display the current server time on the info panel. Hopefully this should make things a little clearer for users. I'll upload the change over the weekend.

SigmaEnigma Mar 15, 2019 8:00pm | Post# 1165

{quote} I've just checked, and Forex.com appear to have a server time of GMT-12 (https://www.myfxbook.com/forex-broker/forex-com/10). The EA checks the trading hours set against the current broker's time. So when our orders opened at 17:00 GMT, your broker's server time would have been 05:00, which would explain why it didn't open an order. I would recommend double-checking your broker's offset and using the new Broker Server Offset parameter to obtain the correct time window. e.g. if your broker offset really is -12, then set the start...
Thank you for the information. Everywhere I looked said Forex.com was GMT with no offset, that site is the first to say they're GMT-12 so I figured 7 and 19 with no offset were the correct settings. Now it makes sense as to why the trade didn't execute. I appreciate the investigation.

PaulDV Mar 15, 2019 8:02pm | Post# 1166

{quote} Thank you for the information. Everywhere I looked said Forex.com was GMT with no offset, that site is the first to say they're GMT-12 so I figured 7 and 19 with no offset were the correct settings. Now it makes sense as to why the trade didn't execute. I appreciate the investigation.
It's probably worth giving them a call anyway just to be sure, if on;y to see if they're on DST, in which case it'll be -11 instead.

samfor Mar 15, 2019 10:41pm | Post# 1167

{quote} Because I was using PaulDV's EA, I was sticking to the 7 and 19 time-frame he set sp1031ecifically for that EA which are the default values for GMT servers. I think those numbers are right, I'm just confused as to why others got a buy signal when on my end the EA was out of trading hours. So did it work as intended then? Or are 7 and 19 not the correct values I should be using? I'm using Forex.com as my broker.
I use Forex.com broker with PDV_1031, settings 7 and 19 broker offset 0. It made a buy for me.

Dheck Mar 15, 2019 10:42pm | Post# 1168

{quote} Looks like you are out of trading hours. We all got a trade. Some of us are on broker with +3 !and some on +2
hai, cdsaa .. can you help backtest this EA ULTRON PDV 1031 with a set
stop loss: 80
take profit: 100
Fixed Lot
0.01
trailing start: 18
step : 2

I think the settings are pretty good but I am doubtful because 99.90% have not ticked the data Thank you

cdsaa Mar 16, 2019 12:19am | Post# 1169

2 Attachment(s)
{quote} In LITE_3 there is no risk percent, I have already explained this, it uses a different system for calculating the initial lot, even in the settings it is called Risk Ratio for example: balance 3000, RR 10 - the initial lot will be 0.90, balance 1000 , RR 10 - the initial lot will be 0.30 and so on) but these values may differ slightly there is a dependency on account leverage (AccountBalance() * AccountLeverage() * risk_percent * 0.01) I was interested in your picture with the test LITE_3. How did you manage to get $ 1,000,000 from $ 400...

@Feex,

I have got the reason why it is able to make $400 to $1000000 in 9 months It is not the settings.

It is the Leverage and lot size calculations. As consecutive losses on this is 1 at any time, it is actually using your high lot calculation using leverage and making big profit. I am testing it with 500 leverage. Will post the result file.

Attachment 3270125

With those high numbers, you can always withdraw your profit and let the EA run its own course.

It is doing in real the same, now i understand, why lot calculation is so high on my real account on Ulite_3 with 10%.

Anyway..

Point I outlined on 104 upgrade list- lot calculation, based on advice by one of the users here, will play significant role in returns.

We also need to think about keeping our SL & TP hidden. These are heavy pips.
Ulite_500Leverage_DefaultSettings.zip

shadikalaji Mar 16, 2019 12:45am | Post# 1170

Hi
Please only traders who have experince are invited to reply..
After years of discussion on FF about not using a stop loss and replacing it with dollar cost averaging or keeping the loosing trade like inventory untill the price comes back again, who would agree?
And of those who agree, who can share their own real results

fareed51 Mar 16, 2019 1:42am | Post# 1171

{quote} Looks like you are out of trading hours. We all got a trade. Some of us are on broker with +3 !and some on +2
yes IcMarket +3..

paroucia Mar 16, 2019 2:48am | Post# 1172

1 Attachment(s)
Hi Guys
I'm trying to backtesting but this is what happend:
Click to Enlarge

Name: Screenshot2.png
Size: 18 KB

jamco388 Mar 16, 2019 3:05am | Post# 1173

Hi Guys I'm trying to backtesting but this is what happend: {image}
Amazing..... Lol

cdsaa Mar 16, 2019 4:16am | Post# 1174

1 Attachment(s)
{quote} hai, cdsaa .. can you help backtest this EA ULTRON PDV 1031 with a set stop loss: 80 take profit: 100 Fixed Lot 0.01 trailing start: 18 step : 2 I think the settings are pretty good but I am doubtful because 99.90% have not ticked the data Thank you

here your results.

Click to Enlarge

Name: Screenshot1.png
Size: 204 KB

Feex Mar 16, 2019 4:26am | Post# 1175

{quote} @Feex, I have got the reason why it is able to make $400 to $1000000 in 9 months It is not the settings. It is the Leverage and lot size calculations. As consecutive losses on this is 1 at any time, it is actually using your high lot calculation using leverage and making big profit. I am testing it with 500 leverage. Will post the result file. {image} With those high numbers, you can always withdraw your profit and let the EA run its own course. It is doing in real the same, now i understand, why lot calculation is so high on my real account...
hi cdsaa Thanks again for the test with high quality (99%) - this is closer to reality)
I used 1: 300 leverage, so I could not get such results. But I saw the dynamics on the chart, so I was interested in the indicator of continuous losses, and of course the drawdown. What I do not like about LITE_3 - a tangible accumulation of profit begins to occur only towards the end of the year. And I do not know how to make the accumulation of profits smoother ... (

P.S. What do you mean when you say to hide SL and TP? Hide them from a broker - make it virtual, or something else? Sorry, English is not my original language)

PaulDV Mar 16, 2019 4:48am | Post# 1176

{quote} Thank you for the information. Everywhere I looked said Forex.com was GMT with no offset, that site is the first to say they're GMT-12 so I figured 7 and 19 with no offset were the correct settings. Now it makes sense as to why the trade didn't execute. I appreciate the investigation.
Actually, I may owe you an apology - the myfxbook page I linked to may not be accurate, and I think it's worth you contacting your broker to make sure.

Now for the apology...I think there may be something wrong with the time filter code!

I copied Pawel's original code to begin with, and amended the calculation used, but I didn't think to check the MQL4 functions used. The time filter uses the function "Hour()" to check the trading times, but the MQL4 documentation says that this is the last known server time "when the program starts and is not updated while the program is running".

If this is correct, then it is a lottery as to WHEN you start the EA, as that will be the hour that the program thinks it is...all the time! So, if you start the EA out of trading hours, it will always be out of trading hours!

The problem with this is that this doesn't work this way in the strategy tester, so for a developer who doesn't know this, the EA LOOKS like it's working when tested but when run for real, it doesn't.

I've amended the time filter to use the correct functions "TimeHour(TimeCurrent())", which IS updated on every call, and will post the new version later today.

Feex Mar 16, 2019 4:55am | Post# 1177

{quote} I've just checked, and Forex.com appear to have a server time of GMT-12 (https://www.myfxbook.com/forex-broker/forex-com/10). The EA checks the trading hours set against the current broker's time. So when our orders opened at 17:00 GMT, your broker's server time would have been 05:00, which would explain why it didn't open an order. I would recommend double-checking your broker's offset and using the new Broker Server Offset parameter to obtain the correct time window. e.g. if your broker offset really is -12, then set the start...
I also encountered such questions, so in LITE_3 I displayed the local time and server time so that it was immediately visible and clear what time settings to use)

cdsaa Mar 16, 2019 5:11am | Post# 1178

Point i mentioned on Virtual SL & TP, it is mainly to hide from Broker. ECN brokers, we will not have problem.

That is on the wish list.

Now lets spend time on improving on our losers.

1. Point i have noticed when analysing losers is, they are stagnating for more than 8-10 hrs before falling to death.
2. May be we bring in a parameter for input hours- say i can input 6 or 8 or 10

and EA checks how much is difference in negative. If negative is more than 5pips, it closes the trade, if it is positive, let it ride. that way we can test and see if it minimizes the losers.

Our EA calculation of Moving Averages Gap between fast and slow, is accurate from August. because of Big Swings. before that it is just ok, with still winners more than 60%. so we may have a good one even then, if we can minimise losers.


On winning trades, within few hours, it is going into first modification on trailing stop. So, if we dont have positive pips we can close such trades. lets see @feex or @PaulDv can put this into working.

PaulDV Mar 16, 2019 5:14am | Post# 1179

{quote}We also need to think about keeping our SL & TP hidden. These are heavy pips. {file}
I've coded hidden SL and TP before and they are a right pain in the arse to do, especially with multiple orders and moving stops. The only real reason to use them is if you don't trust your broker and think they may be stop hunting...in which case, you are better off finding a broker you DO trust.

To be honest, your comment raises several things in my mind that I think are worth pointing out:

  1. You shouldn't keep that amount of money in your trading account:

    1. Black Swan events do happen, and they can wipe out your entire account!
    2. Brokers can go bust, and they will take your funds down with them.
    3. Regulation only protects about the first 20k - anything more than that will be lost.

  2. You should be using a true ECN broker:

    1. The broker makes its money solely from commissions and swaps and does NOT make money when you lose, so there's no conflict of interest.

  3. The test results are not accurate once you get to that size, because orders of that size may not get filled, either on open or on closure, due to lack of liquidity.

    1. If you use a broker that "guarantees" your orders will be filled, I can assure you that they are crooks: they take the other side of your order (basically they are betting AGAINST you), and make money when you lose, and lose money when you win!
      They have a vested interest in making sure you lose!
      These brokers are NOT ECN brokers, and most likely are not regulated. So if your current broker "guarantees" your orders will be filled, you need to change brokers NOW.

When your account gets to being several hundred thousand in size (if you get that lucky!) then you should employ an 80/20 rule: pull out 80% of your capital and save it in a non-broker account, and then gear your EA to trade the remaining 20% as though it was the 100%. You will probably need high leverage for this though, and may have to play around with the percentages to make this work (maybe 50/50 will be enough for you). But the idea is that you protect your main capital, and if your trading account does suffer a wipeout, you still have capital to start again.

Of course, the choice is yours, but this is what I aim to be doing.

If you're wondering about what good broker to use, then, if you're not in the USA, I can thoroughly recommend my broker IC Markets. They are regulated, true ECN leverage of up to 500:1, and have some of the smallest spreads around. I've been with them for well over a year and have had no problems with withdrawing funds. Check them out.


Feex Mar 16, 2019 5:15am | Post# 1180

{quote} Actually, I may owe you an apology - the myfxbook page I linked to may not be accurate, and I think it's worth you contacting your broker to make sure. Now for the apology...I think there may be something wrong with the time filter code! I copied Pawel's original code to begin with, and amended the calculation used, but I didn't think to check the MQL4 functions used. The time filter uses the function "Hour()" to check the trading times, but the MQL4 documentation says that this is the last known server time "when the program starts and...
Yes, in the original version the clock does not work. For example, if you need to set the start time to 23 and finish 6, so that the adviser trades from 23 to 6 in the morning, this does not work in OP1. But OP1 is just the basis of the system itself, without additional fixes. So do not worry)


© Forex Factory