Multi 10:4 Trader EA

[slipshod]

Recently you added the trailpercent exit. This method of exit already locked in many profits for us and is a good addition to the strategy.

I'm wondering why you opted for using the basket method. I never understood the use of using baskets and nobody has ever been able to show me the advantage of using baskets.

You only should manage your account, that is your basket!

If I would say to you: "The AUDCAD is up 5%, so buy GBPCHF, NZDUSD and EURJPY", would you think is logical? probably not. So why would you accept this on the exit side?

I hope you don't mind my jumping in - I'm a little confused by what you're asking. You say the trailpercent is a good addition and has locked in profits, but then question why its been added & can't see the advantage? Isn't that a little contradictory? Also the example you provide is relating to entries, but the EA only basket-manages trades that are already running, with entries judged on individual pairs alone.

Let's say you have a basket that is performing good. It has 5 pairs in the basket and 4 are each up 4 percent and rising. (so 16 percent total) One stock is going down 5 percent. What would you like to happen? Close the whole basket, or just the one losing pair?

In a case like you describe here, the falling trade could potentially drop the overall profit and hit the trailing percentage, which would indeed close the whole basket. It would depend on how sharply that one trade's falling compared to the four others are rising, how tight the trailingpercent stop is, and where the individual losing trade's stop-loss is - if it stopped out the problem would be solved.

Truth is no-one knows what the market will do next. That trade could turn around and run into profit while the others stall and reverse. At some point you have to make a decision to take money off the table and await new entries, and the advantage of the trailpercent is that you know for sure that a proportion of profit gained is going to be kept - not lost because the trader lacks judgement & lets winners turn to losers (or is asleep while the market moves overnight).

[SWG123]

You could try GridMethod set to 1.

Andy

Nope, that doesn't work either, still not picking up the second trade.

[SWG123]

My "missing trade" problem could well be Empty4-related. I'm noticing some non-standard behaviour in Terminal. Normally, when you hover the mouse cursor over a trade, the ticket number, trade comment and magic number fields appear and remain visible for about five seconds. With the trade that doesn't appear in the dashboard, however, the fields appear briefly but then dissolve almost immediately.

[linuspoon]

Linus,

could you try the following and give me some feedback about the error:

In the globals declaration section:

Code: Select all

int    wide = 9;
int    wide2 = 5;      // insert this line
string headerString[] = { "PAIR:", "CHART:", "TREND:", "AO:", "STOCH:", "PRO:", "STOP-TRADES:", "" , "  OPEN-TRADES:"  , "                CORRELATED PAIRS:"};
string headerString2[] = { "LOTS:","PIPS:", "CASH:", "B-EVEN:", "ASK:"};

and in function CreateGrid:

Code: Select all

    for (int x = 0; x < wide + 1; x++)
    {
        sObj = "textheader" + x;
        LabelCreate(sObj, corner, gridCol1[x] + offsetX2, offsetY - 20);
        LabelSetText(sObj, headerStr[x], fontSize, headerfont, headertextColor);
    }

    for (x = 0; x < wide2 + 1; x++)       // change from wide to wide2
    {
        sObj = "header2" + x;
        LabelCreate(sObj, corner, gridCol2[x] + offsetX2, offsetY - 20);
        LabelSetText(sObj, headerStr2[x], fontSize, headerfont, headertextColor);
    }

Background:
Andy used variable wide for his loops over both arrays headerstring and headerstring2. headerstring2 actually has only 5 elements. I adopted this method without thinking about possible consequences (copy and paste ... ). IMO this should lead to more runtime errors than we are actually facing. Why only Linus?

@Andy: maybe it would be better to change this in DisplayGrid() also.

@Linus: hoping for your feedback about this hack.

simplex

Hi simplex,

I have made the changes and recompiled the mq4 but the error is still there with the new compiled EA.

While trying to debug on my own, I've also made the changes in the DisplayGrid() function in the for loop of headerString2. I noticed that there is another footerString section in the DisplayGrid() function which uses the variable wide as well although this only contains 2 elements. So I've also added another variable wide3 and changed the for loop for that to use wide3. After all these changes, the error is still there after recompiling.

Any other ideas?

Linus

[MrLong]

Hi Andy,

I'm still following your efforts on this complex EA.

Recently you added the trailpercent exit. This method of exit already locked in many profits for us and is a good addition to the strategy.

I'm wondering why you opted for using the basket method. I never understood the use of using baskets and nobody has ever been able to show me the advantage of using baskets.

You only should manage your account, that is your basket!

If I would say to you: "The AUDCAD is up 5%, so buy GBPCHF, NZDUSD and EURJPY", would you think is logical? probably not. So why would you accept this on the exit side?

Let's say you have a basket that is performing good. It has 5 pairs in the basket and 4 are each up 4 percent and rising. (so 16 percent total) One stock is going down 5 percent. What would you like to happen? Close the whole basket, or just the one losing pair?

Just my thought.

Hi,

I see basket trading as giving the trader an edge, where as a bookie more or less always wins, why does he win?

In my opinion, it's because there are, let's say eight horses in a race, the guy goes into the bookies and places a bet, he thinks horse number 2 will win, so he has 1 chance of winning, the bookie has 7 chances of winning. So if we have 8 trades rolling at the same time, I think with the 10.4 strategy we have more chance of winning than losing if we cash the basket in for a profit.

On the other hand, you are right in what you are saying, why cut winning trades?
How long will they be winning trades?

I'm always looking for ideas that can give us an edge, so if you have any that we could implement
I'd be please have a look at the ideas.

You can switch off the basket settings by entering a 0.

Many Thanks
Andy

[rosst]

I am getting this message over and over: Order, EURGBP,Short closed by BE function
but no orders have been closed or even look like it applies...
I have closed and restarted but it still goes on....

[MrLong]

Hi Linus,

I never faced this error - neither on my development machine nor on test.

Can you provide ANY information about the circumstances this error occurs? Only on EA startup? On every tick? After trade open? After trade close? ... etc. The more detail you can provide the better. Already did a code review, found nothing ...

Cheers, simplex

This error only occurs in one of my demo Empty4 but not on another one. The error occurs once the EA starts. On every tick (may not be every tick but there are a number of errors within 1 min interval), the same error will appear in the Expert tab one line after another continuously until either I stopped the EA or change the GridMethod to 1 or if I set showDisplay to false.

Linus

Linus,

could you try the following and give me some feedback about the error:

In the globals declaration section:

Code: Select all

int    wide = 9;
int    wide2 = 5;      // insert this line
string headerString[] = { "PAIR:", "CHART:", "TREND:", "AO:", "STOCH:", "PRO:", "STOP-TRADES:", "" , "  OPEN-TRADES:"  , "                CORRELATED PAIRS:"};
string headerString2[] = { "LOTS:","PIPS:", "CASH:", "B-EVEN:", "ASK:"};

and in function CreateGrid:

Code: Select all

    for (int x = 0; x < wide + 1; x++)
    {
        sObj = "textheader" + x;
        LabelCreate(sObj, corner, gridCol1[x] + offsetX2, offsetY - 20);
        LabelSetText(sObj, headerStr[x], fontSize, headerfont, headertextColor);
    }

    for (x = 0; x < wide2 + 1; x++)       // change from wide to wide2
    {
        sObj = "header2" + x;
        LabelCreate(sObj, corner, gridCol2[x] + offsetX2, offsetY - 20);
        LabelSetText(sObj, headerStr2[x], fontSize, headerfont, headertextColor);
    }

Background:
Andy used variable wide for his loops over both arrays headerstring and headerstring2. headerstring2 actually has only 5 elements. I adopted this method without thinking about possible consequences (copy and paste ... ). IMO this should lead to more runtime errors than we are actually facing. Why only Linus?

@Andy: maybe it would be better to change this in DisplayGrid() also.

@Linus: hoping for your feedback about this hack.

simplex

Hi Jurgen,

If it's an improvement, change whatever you feel would benefit the EA..

Thanks
Andy

[fx800]

I loaded v2 yesterday OOTB onto an FXCM demo and it had a very bad day - over 5% down. IS this
a) usual variance and I just turned on wrong day or
b) an error somewhere in my set up ie did anyone else have a negative day yesterday?

Hi Doc, I've got it running on several demo accounts, one of which is linked in my sig below. Yesterday wasn't great, and so far today's been no picnic either - but on the other hand there's been days where it's gone gangbusters. I haven't experienced 8% DD yet though, but as each Empty4 instance seems to take different trades its certainly possible - and that's the one thing I don't feel comfortable about with this EA, the fact that the trades taken seem to vary so much. For instance, about an hour ago my Cowboy IBFX demo went long on USDCHF at 0.948 and is now 30 pips in profit, whereas my other demos only have buy stops higher up at 0.952.

Anyway, my guess is you started on a bad day. I'm expecting this will need weeks/months of solid forward testing to get a real feel for whether or not it'll be consistently profitable.

To-day is a clear illustration that you have to actively manage the ea and not use it as set and forget.

If you happened to visit Bob's 10.4 thread yesterday evening at around 10.43pm London time, you would have seen this
http://www.stevehopwoodforex.com/phpBB3 ... start=2710

The indicators means nothing, he said. Since the ea is based on indicators, you would not expect it to work when there is a major fundamental news coming out imminently.

Based on his view, I switched off JPY pairs, and placed pending orders manually for buys for xxxjpy pairs. I am not as brave as Bob and only did four pairs AJ, NJ, GJ and CADJPY out of a possible seven and got several hundred pips. All this while I was asleep and set my basket tp to 10%. For some reason, the basket tp did not work (even though the laptop was on) and ended with a 22% profit which I closed manually at 7.30am.

What a night !

Hope some of you were in on the trades.

peter

[MrLong]

I loaded v2 yesterday OOTB onto an FXCM demo and it had a very bad day - over 5% down. IS this
a) usual variance and I just turned on wrong day or
b) an error somewhere in my set up ie did anyone else have a negative day yesterday?

Hi Doc, I've got it running on several demo accounts, one of which is linked in my sig below. Yesterday wasn't great, and so far today's been no picnic either - but on the other hand there's been days where it's gone gangbusters. I haven't experienced 8% DD yet though, but as each Empty4 instance seems to take different trades its certainly possible - and that's the one thing I don't feel comfortable about with this EA, the fact that the trades taken seem to vary so much. For instance, about an hour ago my Cowboy IBFX demo went long on USDCHF at 0.948 and is now 30 pips in profit, whereas my other demos only have buy stops higher up at 0.952.

Anyway, my guess is you started on a bad day. I'm expecting this will need weeks/months of solid forward testing to get a real feel for whether or not it'll be consistently profitable.

Hi Guyz,

A long time ago I discovered the data given to us on a demo account is
different from their live data feeds, and different from broker to
Broker, click on the 4th speed button from the left, the data window
Will appear, you can compare the brokers prices for MA,s and pivots.

I haven't compared them in recent years, but they were always
different, a few months ago I had the same EA loaded on Globle and Alpari
none of them took the same trades.

Andy

[MrLong]

You could try GridMethod set to 1.

Andy

Nope, that doesn't work either, still not picking up the second trade.

That's why I entered the magic numbers in the comment, to try debug, can you give me a screen shot of the terminal.

Andy