Discussion - page 67
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
Download excel files from the first post of this thread https://www.mql5.com/en/forum/176044 and you will see which one is the best. But please note that i am testing not all EA posted for elite section so may be there are some EAs which Inever tested sorry.
Some people asked me by PMs the following:
- "i am new to forex so where to read?"
- "i want to learn programming but do not have much time for that so where to read?"
You can read here https://www.mql5.com/en/forum
help please
I have an internet connection issue that is not easily solved because I live in a rural area that doesn't have high speed internet and at times this interferes with my EA's trading performance. What I would like to do is set myself up with some kind of system to resolve this issue. What I was thinking was some sort of shared server or dedicated server or I don't know what and thats why I'm asking you guys to guide me with what I need.
puzzling Steinitz HAS backtest problem
Hello,
I am posting this here as well in hopes that some more experienced MT4 experts will be able to help me solve a backtest puzzle. I am backtesting the HAS MTF Stenintz v 2.63 (default settings) with 90% modelling quality and I get good stats up to the last trade that loses all the profits. I call this the "spike of death"! In the attached backtest report you will see this behavior. Here is the the last large losing trade from the report:
330 2007.06.22 13:00 buy 14 0.10 124.07 0.00 0.00
331 2007.06.22 13:00 modify 14 0.10 124.07 0.00 124.37
332 2007.11.26 23:59 close at stop 14 0.10 107.36 0.00 124.37 -1372.00 9143.79
This says: line #330 on 2007.06.22 at 13:00 a buy is initiated for 0.10 lots and trade #14 is entered at 127.07. Line #31 is a TP mod to change TP 0.0 to 124.37 with the datestamp of 2007.06.22 13:00. Line #332 comes in at 2007.11.26 time 23:59. Notice it is 6 MONTHS later from when the trade was opened!! It closes this trade #14 at 107.36 for a WHOPPING LOSS of 1671 pips. There were no trades between 06.22 and 11.26 since there was a trade active which prevented another trade being opened.
Steinitz states that this is because of the backtester and I say it is because the EA fails to close this trade and over few months it turns into a big loser only to be closed as the last trade by the backtester. What do you guys think? Thanks for your help.
no kidding. It has no stop loss. It is a only a matter of time before it loses entire account. This is a mathmatical certainty of occuring.
Hello,
I am posting this here as well in hopes that some more experienced MT4 experts will be able to help me solve a backtest puzzle. I am backtesting the HAS MTF Stenintz v 2.63 (default settings) with 90% modelling quality and I get good stats up to the last trade that loses all the profits. I call this the "spike of death"! In the attached backtest report you will see this behavior. Here is the the last large losing trade from the report:
330 2007.06.22 13:00 buy 14 0.10 124.07 0.00 0.00
331 2007.06.22 13:00 modify 14 0.10 124.07 0.00 124.37
332 2007.11.26 23:59 close at stop 14 0.10 107.36 0.00 124.37 -1372.00 9143.79
This says: line #330 on 2007.06.22 at 13:00 a buy is initiated for 0.10 lots and trade #14 is entered at 127.07. Line #31 is a TP mod to change TP 0.0 to 124.37 with the datestamp of 2007.06.22 13:00. Line #332 comes in at 2007.11.26 time 23:59. Notice it is 6 MONTHS later from when the trade was opened!! It closes this trade #14 at 107.36 for a WHOPPING LOSS of 1671 pips. There were no trades between 06.22 and 11.26 since there was a trade active which prevented another trade being opened.
Steinitz states that this is because of the backtester and I say it is because the EA fails to close this trade and over few months it turns into a big loser only to be closed as the last trade by the backtester. What do you guys think? Thanks for your help.no kidding. It has no stop loss. It is a only a matter of time before it loses entire account. This is a mathmatical certainty of occuring.
Thanks for answering. It seems that the losing trade did not have a stop loss. However looking at the other trades, they are modified with SL after they are opened. Looks like the SL for this trade was not managed for some reason. Could this be caused by the backtester or it it an EA issue in your opinion?
From the backtest report you can see that trades have no Initial Stop and Trailing Stop is activated when trade is in profit.
So it's not the problem of the Tester, it's a problem of EA.
Thanks for answering. It seems that the losing trade did not have a stop loss. However looking at the other trades, they are modified with SL after they are opened. Looks like the SL for this trade was not managed for some reason. Could this be caused by the backtester or it it an EA issue in your opinion?
Hi fxspeedster,
MTF EAs are not easy for backtesting. The problem is the following: we must know how it was coded. We are having few MTF EAs in elite section and we always can see inside the code to know how it was coded.
Why we need to know how it was coded?
Because if it was coded on open bar so backtesting is not reliable.
Example: Firebird EA was coded on open bar and backtesting is very different from forward testing. MaChannel EA was coded on high/low of the close bar and forward testing performance is more good than backtesting.
In case of Steinitz we may have two cases:
- it was coded on close bar. But in this case the chart (manual trading system) will not be the same with MTF EA. Can you imagine that current signal will be filtered by monthly chart on close bar? Monthly chart on close bar is 1 month ago! Because close bar is previous bar and previous bar for MN1 chart is last month.
- so most probably it was coded on open bar. In this case backtestibng will not be the same with forward testing in most cases: forward testing should be better than backtesting, or backtesting will be better than forward testing.
Look at this image. It is not Steinitz. It is market condition evaluation from here:
- AbsoluteStrengthMarket indicators: indicators to analyze the market condition. Read posts starting from this page to this one.
- indicators and template to estimate the market condition for D1 timeframe is here.
Do you see AbsoluteStrengthMarket indicator on the image?
- Previous small square on Current is previous bar. It is close bar.
- Previous small square on W1 is open bar (in most cases). D1 chart.
- Previous small square on MN (D1 chart) is open bar on MN (in most cases). Close bar on MN is many small squareas back and it is last month.
So if you are trading MTF manual trading system you are using some MTF indicators as well on open bar.
Thus, if Steinitz EA was coded onj open bar by some part of the code so backtesting is not reliable. Forward testing only.
And as to your close at stop order so it is the same as you close the order manually. For example you are testing some EA and decided to close open order manually now. And posting the statement now. Order was closed because backtesting was finished.
Thus basicly it is the drawdown.
To say the true almost all MTF EAs are having big drawdown.
As to Steinitz EA so ... may be it is good EA ... I have no idea because I do not have it and never tested it and I do not know how it was coded: commercial sellers do not co-operate with me with any way.
We are having our elite section Eas which are much better and less insinuated as Steinitz EA so sorry don't want to discuss any commercial EA in elite section anymore.
Time Based Trading
Hello all,
I am working on an EA, in which I need to trade at specific times each day. My problem, I don't know how to code the function of
specific trading times.
Any suggestions??????
Thank in advance.
jayborde
Timefilter:
- how to set EA's time in timefilter is on this post and this page.
- how to code timefilter in EAs (codes) is here.
- "Non-Trading Hours" on the screen fixing: example code; now it displays "Trading Hours" during trading hours and "Non-trading Hours" during non-trading hours. Thanks Locutus.