Hello Aha,
I am testing all 19 pairs available from IBFX. I have also found that in the morning (my local time is GMT -06 (US CET) ) Mouteki EA often has profitable trades opened. But if I don't close them manually many winners turn into losers later. Could you create an option to close all trades when total profit reaches user defined equity level (for example $300)?
// Ver 0.4 Added parameters to control the mark hours the EA is allowed
// to enter a trade.
The idea behind this is that, the success rate of a TL break trade is directly related to the momentum behind the move that breaks the TL. Apparently most of the trades were placed in EUR open and US open and therefore the market force in those hours should be stronger and more determined. Once the price breaks the trend line it's more likely to go forward then to go back. While in other hours most of the time the market is more tame and less directional --- which tends to produce false break outs. In v0.4 we can only allow EA to open trades in US and EURO marketing hours, to reduce bad trades.
I got this idea because in the recent a couple of days, Mouteki was winning big in the morning, but return most of the profits to the market in the afternoon. Closing trades in the morning alone is not enough because the EA will open trades based on false breakout later. So I though it would be nice if we can forbid the EA to trade in those hours.
This is just my observation, let's test it together and see if it's a valid filter.
Ignored
Aha
Thanks a lot for your continued efforts to improve the EA, just one question: should the computer and system be running on a 24h basis for optimal results?
AHA,
Thank you for all your work and creating this thread. I have the Bago AHA zip downloaded to my MT4 platform and its on my indicators list. So if I understand you, each alert will send me an e-mail. My question is, how do I modify it to send me a text message at the same time? thanks much
Thanks for confirming that. What you have suggested is not difficult to implement, but I prefer wait a little longer to get more data. Besides, I am pretty busy with other projects right now.
BTW, anyone who is interested please feel free to improve all these EAs in your way.
aha
Quoting tomstaufer
Disliked
Hello Aha,
I am testing all 19 pairs available from IBFX. I have also found that in the morning (my local time is GMT -06 (US CET) ) Mouteki EA often has profitable trades opened. But if I don't close them manually many winners turn into losers later. Could you create an option to close all trades when total profit reaches user defined equity level (for example $300)?
I still don't understand why checking two bars would miss the signal. Bar 0 is not closed yet. The difference is, you check it before closing and I check it after closing. So you would enter earlier but I still get the signal at the beginning of the next bar.
Quoting aha
Disliked
About OR or AND, here is quote for the filter:
[QUOTE]They then introduce “TD Line Qualifiers”, with four different parameters to qualify a trendline breakout. In the article they state, “If any of the four qualifiers are true, the trendline break is valid.” Following are the four qualifiers for a supply line breakout and (hopefully) accurate examples of each one[/QUOTE].
About the number of bars to check, here is the reason:
There were situations when there is a new support or resistance point formed at the 3rd bar therefore a new TL just formed. At the time if you only check the recent two bars you won't find the price have broken the new trend line because the bar 0 and 1 may be both below or above the new TL, there you will miss a signal.
Let's say the previous support point was bar 10 (irrelevant here), and a new support point formed at bar 3 when the bar 0 just opened. Our EA recognizes the new support point immediately after bar 0's open, so a new bottom TL is formed.
To check if the TL is broken, your EA will check if bar 1's open is above the TL and bar 0's open is below the TL. If I didn't remember wrong your EA first calculates where the TL intersect the bar 0 (or its extension line), then compare the value with bar 0 and 1's open price.
My EA will calculate where the TL intersect the bar 0, 1, 2 and compare the three values with bar 0, 1, 2's opening price in respectively. Do you see the difference now?
There are two possible problems here for your EA (if I didn't remember wrong): firstly, though bar 1 's close has to be higher than bar 3 's close to make bar 3 a valid support point, bar 1 's open may be lower than the new TL if the TL is steep enough (if bar 3 is short, the TL don't even need to be too steep). Because both bar 0 and bar 1's open are below TL, your EA won't recognize this TL break so takes no action. Secondly, even when bar 1's open is above the TL, it may be lower than where the TL intersects bar 0 's extension line (the TL is tilted upward), so at that situation your EA won't be able to detect the TL break too.
I didn't think this through in the beginning too. But I got many feedbacks pointing out that I was missing some trades so I got it figure out. This is why I think testing is really important for EA's quality.
Quoting GP2X
Disliked
I still don't understand why checking two bars would miss the signal. Bar 0 is not closed yet. The difference is, you check it before closing and I check it after closing. So you would enter earlier but I still get the signal at the beginning of the next bar.
Let's say the previous support point was bar 10 (irrelevant here), and a new support point formed at bar 3 when the bar 0 just opened. Our EA recognizes the new support point immediately after bar 0's open, so a new bottom TL is formed.
To check if the TL is broken, your EA will check if bar 1's open is above the TL and bar 0's open is below the TL. If I didn't remember wrong your EA first calculates where the TL intersect the bar 0 (or its extension line), then compare the value with bar 0 and 1's open price.
My EA will calculate where the TL intersect the bar 0, 1, 2 and compare the three values with bar 0, 1, 2's opening price in respectively. Do you see the difference now?
There are two possible problems here for your EA (if I didn't remember wrong): firstly, though bar 1 's close has to be higher than bar 3 's close to make bar 3 a valid support point, bar 1 's open may be lower than the new TL if the TL is steep enough (if bar 3 is short, the TL don't even need to be too steep). Because both bar 0 and bar 1's open are below TL, your EA won't recognize this TL break so takes no action. Secondly, even when bar 1's open is above the TL, it may be lower than where the TL intersects bar 0 's extension line (the TL is tilted upward), so at that situation your EA won't be able to detect the TL break too.
I didn't think this through in the beginning too. But I got many feedbacks pointing out that I was missing some trades so I got it figure out. This is why I think testing is really important for EA's quality.
Aha, I tried your latest version but it still generates too many fake signals. Since this is a coutertrend method, could you add more filters? Like divergence, you can use MACD or CCI. Another filter might be only trade when the break happens arround fib lines(using the vegas 4 hours method).
I think whether this method would work really depends on how people draw the trend lines, so it's not a mechanical method.
Yeah I thought about that, and even developed a MACD divergency indicator to see how reliable it is. But the result confirmed that divergency has less than 50% success rate in signalling trend reversal, which was stated here: http://www.investopedia.com/articles...ACDdiverge.asp
About fibo lines, I just come cross a method called Murrey Math, which draws lines to show the strong/weak support/resistance. I haven't found enough free materials to read yet. A software based on it was sold for over $2000+ and the manual was for $70+.
Anyone have any info about it?
BTW, as someone pointed out before, Mouteki system is a variation of Thomas Demark's entry method, hardly original or complete.
Quoting GP2X
Disliked
Aha, I tried your latest version but it still generates too many fake signals. Since this is a coutertrend method, could you add more filters? Like divergence, you can use MACD or CCI. Another filter might be only trade when the break happens arround fib lines(using the vegas 4 hours method).
I think whether this method would work really depends on how people draw the trend lines, so it's not a mechanical method.
About fibo lines, I just come cross a method called Murrey Math, which draws lines to show the strong/weak support/resistance. I haven't found enough free materials to read yet. A software based on it was sold for over $2000+ and the manual was for $70+.
