Flyer! EA - New Thread - Part 2

Hi,

First, thanks for all the hard work. I'm impressed by your programming skills.

As to future features, I've often thought that any EA could generally benefit by having a different set of parameters for buy vs. sell, considering the differential of wins in each direction. It's probably a lot more work though, so I'll understand if you put that idea on the back burner.

Rob

It is a worthwhile attempt to look at the differences. Although, in much of the backtesting I did the ratio of buys vs. sells and the win/loss ratio seemed fairly equal. In fact, I took great notice in looking at any significant discrepancies between the two. Most longs vs. shorts being unequal during a certain timeframe of several months can usually be attributed to the trend going in that direction.

But nevertheless, it is still worthwhile because there are certain pairs (primarily the JPY pairs) that at this time have a tendency to go up more than they go down.

Thanks for the suggestion.

GF

Could a few that are using this EA post results maybe? I just tried the EA and it opened 2 winning trades. I was just wondering how some are doing using this EA. If it is working great I might try it later live. Just wondering which system is working the best is all. Maybe you could also post individual settings .... that work well or the best for you.


Thanks very much,

ut2DaMax

Yes, the percentage of wins is generally fairly even with your EA. I have been developing one where it showed 78% on Long and 85% on Short. By tweaking the S/L and TP points only, the EA's profitability jumped about 40% without changing either of these win ratios.

A worthwhile exercise, although it does seem a bit like curve fitting unless tested on a relatively long data set. Your discussion earlier about the Alpari data and test setups helped. Nice to learn from those more experienced.

Rob

Marci,

There hasn't been any posts for a couple of weeks. Are you still using the Flyer. What results have you been having?

Thanks,

dbcmkx

I typically will close a position manually to lock in the gains since at times you will see a winning position turn into a loser if you let it close for you. Once of the problems with waiting for any type of crosssing before entering or exiting a position is that it is a lagger no matter how you put it.

DB,

Nobody has had any questions recently and I haven't made any updates to the program either. I have been working on some other private projects so I haven't devoted as much time to this one. Although, I am looking to make some improvements soon enough to improve the exit strategy the EA uses. As ET points out, its biggest weakness is allowing good winning trades to be reduced into poor winning trades or worse yet, losers.

GirlFlyer

I can just say that this is probably one of the better EAs in this forum. I must still manually close positions to keep them as winners.

I'm glad that you feel that the EA is worthwhile. I have found that it does generally make a correct entry decision, but often times since the EA uses the M30 timeframe for its entries, the same timeframe does not seem to work well with the exits because of the lag... so, perhaps this may be a clue to a future methodologies of exit strategies for the EA.

GF

Girlflyer

I am not quite sure how to exactly handle the exit rules, I cannot find exiting on lower or higher TF as to which is better than the other. Sometimes I see that either the lower or higher TF will save me, but I am mostly interested in preserving the gains. I guess I should pay more attention to the indicators that may possibly give a hint to exit a position. I had been looking at ADX with the DI+/- as a possible indication. Another possibility is the GMMA [Guppy MMA], I do not know if you know anything about it.

ET

Hi G.F.

May be you are already aware that one distinguished member of this forum has made an EA suitable for exits.
I have not test it, as I am not able technically wise, however you may give it a try.
All the details are in Trading Systems-PipBoxer-SE, Safety exit.
Hope it may help.

Regards

Sotos

I think the exit conditions are also personal preference which makes it even harder to decide on appropriate conditions for an exit. Do we make an EA such that once a certain profit level is hit, we take the profit once it retraces a certain value? Do we exit the trade only when the proper exit signal is made? There are so many choices to decide. I often use the EA for an entry point on manual trades (but even then, I don't trust every entry either) but rather than letting the EA decide the exit, I decide when it is better to get out because while the entry can be appropriate, the price action can often reverse quickly. Last Friday's NFP was a good example of that. Yet, when getting out of trades early, we go against the concept of the system. The system is designed to follow a trend to its maximum potential. The hardest part is determining where that maximum is. Sometimes the maximum is reached and it stays there for some time giving the indicators time to catch up and it can sell out with most of the profit. Other times, it reaches the max, is there for a short time, and begins to reverse maybe not quickly, but reverses and the EA can grab 60-70% of the potential with some diligence. Then you have the times when the reversal is quick and sudden. When that happens, the EA would need to act quickly to preserve any gains at all and more than likely a S/L would have to be used to prevent a position going negative. Anyway, Flyer currently is designed to follow the trend and tries to take full advantage of it. If the price action remains fairly stable on a consistent trend without a lot up/down volatility, it will and can follow a move for probably 2,3, even 400 pips. But how many times do we see that happen? Most of the time things seem to move 50-100 and then stops and consolidates for awhile, retraces some, then maybe goes onwards. These are awkward moments and are hard to program for. So you see the dilemma.

To be frank with everyone, while I developed Flyer with the concept of using it to go along with a system of mine, my own manual trading I do nowadays has developed into a method far from what I have implemented with Flyer and is much more consistently profitable (No, I'm not going to put an EA out on it.... sorry). To hint at it, I pretty much use price action alone without involving myself in complicated indicators and MA's. It happens sometimes that these trades coincide with Flyer's decisions, but more often not. Now, what I may do is take some of the rules I use with my strategy and apply some of it to Flyer for future and see how that works out.

Ok, this is a long post and is more like a commentary/journal entry for Flyer. I hope you find the reading worthwhile.

GirlFlyer

Please see the attached file. Flyer2.0 triggered and closed 3 consecutive GBP/USD trades. Why?

Thanks,

dbcmkx

Marci,

I Have attached a file showing the input settings that I use. Since my previous post regarding this matter Flyer has trggered another 5 GBP/USD trades. I have shut down the ea for this pair. It has not done this for any other pair.

Idos

DBC,

This is a puzzler for me, because the programming logic I have does not allow for the EA to close a losing trade except by Stop/Loss which is done by the broker. You have not made any modifications to the code have you?

I will continue to investigate further, but at this point I don't have an answer for you.

Marci

DBC,

Please send me a copy of your log file from the 'Experts" tab from the time frame of the trades. You can email it to me as well if you like. Just let me know if you email it. Hopefully that will help me to track this down.

Marc

Marci,

Attached is the expert log file you requested. I also emailed it to you.

dbcmkx

Hi again,

The log file was the key.

I figured out your problem... sorry to say it was not literally a problem with the EA. Your broker margin called your trades. It comes down to a simple fact of free margin in your account. The free margin in your account at the time the orders were placed was just enough to cover placing the trade. Then when your trade was made your free margin became nearly zero. Then when the market moved against you, by just 1 or 2 pips, it caused a margin call. They closed out your most recent trade which happened to be the GBP. I could tell this because the EA was also trying to open some trades on the Eur/usd at nearly the same time which showed that the margin was insufficient to cover the trade. Also, the trades on the GBP were stop/loss closed by the broker at values the EA would not/can not set.

Then afterwards, your margin increased enough again to repeat the process. Thus became a vicious cycle.

My suggestion, don't let your EA's (you're using multiple ones I can see) use all of the margin in your account. I doubt if in a live account you would use 100% of your margin because this exact thing would happen there, too.

However, this situation does bring to life a circumstance for the EA which I had not considered before because I've never run into it. There could be a simple code addition in the EA to prevent this kind of occurence by checking to see how much free margin would remain after a trade was placed and if it left too little margin, then it could not allow the trade. I can add this in to an update in the near future.

Generally speaking this is a rare occurence. I actually wonder how many EA's code for this circumstance. With people testing out so many EA's on their system, the possibility of using nearly 100% of margin becomes more feasible I suppose. You may want to check to see if your other EA's prevent trading when the free margin comes close to zero.

I trust this helps you to understand what happened. Free up some margin and you can go back to trading as usual.

Marci

GirlFlyer

I do believe you should code to check for free margin since I had also tested EAs where the trades were just closed or not triggered due to the free margin. I only came across one EA in the past which actually checked for this otherwise the EAs will just drain the account.

Marci,

Please see attached. I read your post, but in this case I don't see how there wasn't enough margin. The EA closed the GBP/USD order 15 seconds after it opened the trade and 40 seconds before it opened the EUR/USD order. There was enough margin to open the 2nd trade and keep it open, and both trades were basically the same size. So if there was enough margin to keep the 2nd trade open, there should have been enough margin to keep the 1st trade open. If the 1st trade was open and the EA tried to open a 2nd trade and found that there was not enough margin, I assume the EA would not open the 2nd trade. However in this case the 1st trade was closed before there was any indication that the EA was trying to open the 2nd trade. The other EA that you saw is an "exit EA" I am testing that only closes trades.

Did I miss something? What confuses me is that it only happens with GBP/USD.

Thanks,

dbcmkx