- #84
- Edited 9:31am Jun 6, 2008 8:56am | Edited 9:31am
Hi again,
I attach here what i coded based on kurka's suppositions that the real price of a currency is the consensus one which can be calculated triangulating with main currencies:
I did it for EURCHF because my broker doesn't have all kurka posted in the first post so i took just 1 pair and tested for that one. SO for euchf we can triangulate this:
EURCHF AUDCHF EURAUD
EURCHF CHFJPY EURJPY
EURCHF GBPCHF EURGBP
EURCHF USDCHF EURUSD
So the price of eurchf in the ea i did is calcualted taking into reference, AUD,JPY,GBP,USD. The consensus is the ma of those.
How it works: if 3 of 4 show the same and consensus also is alligned with them- ea will go and trade. Ea will close when an oposite singal will come. I think kurka did the same but i believe it is not right to close on opposite signal because market can be regulated back slow and we don't get the signal. (ex 2 of 4 go oposite,1 or 4.. after a fiew bars it's back but we haven't got the signal for back)
Anyway the entry i think it's right, what i want is oppinions about a better exit if any interested in this strategy.
About the colors in the top of the screen:
Green - out of std deviation channel: should go up
Blue - it's between std deviation: we don't care
Red - out of std deviation channel: should go down.
First price is real price: (Ask) - second consensus of the next 4.
I used Ask everywhere and added or subtracted (spread+1) pips to check check if price is out of "spread range". Don't know if it's right but seems to show too manny buy signals lol. Mabe (ASK+ BID) / 2 instead? ... The transaction it makes it's only for testing purposes what you should look for is the colors/values in the top-left corner.
Test on EURCHF- m1.
p.s It might have logic errors/codding errors: this was coded in 2 hours and not tested more than 15 minutes
, so please let me know if you find any.
Please let me know what you think, Victor
I attach here what i coded based on kurka's suppositions that the real price of a currency is the consensus one which can be calculated triangulating with main currencies:
I did it for EURCHF because my broker doesn't have all kurka posted in the first post so i took just 1 pair and tested for that one. SO for euchf we can triangulate this:
EURCHF AUDCHF EURAUD
EURCHF CHFJPY EURJPY
EURCHF GBPCHF EURGBP
EURCHF USDCHF EURUSD
So the price of eurchf in the ea i did is calcualted taking into reference, AUD,JPY,GBP,USD. The consensus is the ma of those.
How it works: if 3 of 4 show the same and consensus also is alligned with them- ea will go and trade. Ea will close when an oposite singal will come. I think kurka did the same but i believe it is not right to close on opposite signal because market can be regulated back slow and we don't get the signal. (ex 2 of 4 go oposite,1 or 4.. after a fiew bars it's back but we haven't got the signal for back)
Anyway the entry i think it's right, what i want is oppinions about a better exit if any interested in this strategy.
About the colors in the top of the screen:
Green - out of std deviation channel: should go up
Blue - it's between std deviation: we don't care
Red - out of std deviation channel: should go down.
First price is real price: (Ask) - second consensus of the next 4.
I used Ask everywhere and added or subtracted (spread+1) pips to check check if price is out of "spread range". Don't know if it's right but seems to show too manny buy signals lol. Mabe (ASK+ BID) / 2 instead? ... The transaction it makes it's only for testing purposes what you should look for is the colors/values in the top-left corner.
Test on EURCHF- m1.
p.s It might have logic errors/codding errors: this was coded in 2 hours and not tested more than 15 minutes
, so please let me know if you find any.Please let me know what you think, Victor
- #85
- Jun 6, 2008 10:51am Jun 6, 2008 10:51am
I will put on demo ...
i saw only little set EA , very simple
Thx.
i saw only little set EA , very simple
Thx.
Hi again,
I attach here what i coded based on kurka's suppositions that the real price of a currency is the consensus one which can be calculated triangulating with main currencies:
I did it for EURCHF because my broker doesn't have all kurka posted in the first post so i took just 1 pair and tested for that one. SO for euchf we can triangulate this:
EURCHF AUDCHF EURAUD
EURCHF CHFJPY EURJPY
EURCHF GBPCHF EURGBP
EURCHF USDCHF EURUSD
So the price of eurchf in the ea i did is calcualted taking into reference, AUD,JPY,GBP,USD. The consensus is the ma of those.
How it works: if 3 of 4 show the same and consensus also is alligned with them- ea will go and trade. Ea will close when an oposite singal will come. I think kurka did the same but i believe it is not right to close on opposite signal because market can be regulated back slow and we don't get the signal. (ex 2 of 4 go oposite,1 or 4.. after a fiew bars it's back but we haven't got the signal for back)
Anyway the entry i think it's right, what i want is oppinions about a better exit if any interested in this strategy.
About the colors in the top of the screen:
Green - out of std deviation channel: should go up
Blue - it's between std deviation: we don't care
Red - out of std deviation channel: should go down.
First price is real price: (Ask) - second consensus of the next 4.
I used Ask everywhere and added or subtracted (spread+1) pips to check check if price is out of "spread range". Don't know if it's right but seems to show too manny buy signals lol. Mabe (ASK+ BID) / 2 instead? ... The transaction it makes it's only for testing purposes what you should look for is the colors/values in the top-left corner.
Test on EURCHF- m1.
p.s It might have logic errors/codding errors: this was coded in 2 hours and not tested more than 15 minutes
Please let me know what you think, Victor
- #92
- Edited 4:19pm Jun 6, 2008 4:07pm | Edited 4:19pm
Kurka,
There is also something that i've recently discovered which i don't know how you handle and would like to know, if you can tell me.
If i use the Ask price the triangulations for ask are calculated, and i obtain some values. If i use Bid the values i obtain for the calculated currency are not in the spread zone of the initial calculated values for Ask not even near. So the triangulation for Ask and Bid are much more far between eachothers than real Ask and Bid values.
In this case which one will can be followed? Always i will have a triangulated consensus ask value that is predisposed for showing BUY positions and a bid value that will show more often SELL signals.
Or the negate: If using Ask it would be harder to see Sell signals, and for Bid harder to see BUY signals.
Another problem - technical this time: If the EA runs on currency X and needs the values from currencies A,B,C if A,B,C changes X will not change. If X changes A,B,C values will be read.
So at time T i might have a good signal for Buy or Sell but won't be found/shown by the EA because X was not updated by the metatrader even if A,B,C changed. This way you can fail to open new positions or fail to close when would be the time or close with loss a good position.
RefreshRates doesn't solve the problem because Start is executed only when a new "feed" from currency X was received.
If this cannot be solved i'm thinking of writing my own code that uses the packets metratrader switches on LAN with the broker server. This way it could be solved your other problem of checking some data from another website.
So instead of using a platform i could make my own standalone exe that communicates with broker server and also with other servers for additional informations. Anyway have to see how it performs further, will see...
Attached is a picture of problem 1 if anyone interested to see it.
- check In the top chart is the ASK price triangulated and in the bottom one the BID - check the big differences.
Tjpld: Forgot to tell you, this ea can't be backtested only forwardtested. Backtest anyway can't be done here i would need thick charts very very accurate.The trade is not essential at this point, it was placed there just for test there are bigger problems than that trade
.
Ok, Have a nice weekend, i wait the next week to play with it more. Congrats for the thread/idea kurka, it set me back to work.
There is also something that i've recently discovered which i don't know how you handle and would like to know, if you can tell me.
If i use the Ask price the triangulations for ask are calculated, and i obtain some values. If i use Bid the values i obtain for the calculated currency are not in the spread zone of the initial calculated values for Ask not even near. So the triangulation for Ask and Bid are much more far between eachothers than real Ask and Bid values.
In this case which one will can be followed? Always i will have a triangulated consensus ask value that is predisposed for showing BUY positions and a bid value that will show more often SELL signals.
Or the negate: If using Ask it would be harder to see Sell signals, and for Bid harder to see BUY signals.
Another problem - technical this time: If the EA runs on currency X and needs the values from currencies A,B,C if A,B,C changes X will not change. If X changes A,B,C values will be read.
So at time T i might have a good signal for Buy or Sell but won't be found/shown by the EA because X was not updated by the metatrader even if A,B,C changed. This way you can fail to open new positions or fail to close when would be the time or close with loss a good position.
RefreshRates doesn't solve the problem because Start is executed only when a new "feed" from currency X was received.
If this cannot be solved i'm thinking of writing my own code that uses the packets metratrader switches on LAN with the broker server. This way it could be solved your other problem of checking some data from another website.
So instead of using a platform i could make my own standalone exe that communicates with broker server and also with other servers for additional informations. Anyway have to see how it performs further, will see...
Attached is a picture of problem 1 if anyone interested to see it.
- check In the top chart is the ASK price triangulated and in the bottom one the BID - check the big differences.
Tjpld: Forgot to tell you, this ea can't be backtested only forwardtested. Backtest anyway can't be done here i would need thick charts very very accurate.The trade is not essential at this point, it was placed there just for test there are bigger problems than that trade
.Ok, Have a nice weekend, i wait the next week to play with it more. Congrats for the thread/idea kurka, it set me back to work.
Kurka,
If i use the Ask price the triangulations for ask are calculated, and i obtain some values. If i use Bid the values i obtain for the calculated currency are not in the spread zone of the initial calculated values for Ask not even near. So the triangulation for Ask and Bid are much more far between eachothers than real Ask and Bid values.
In this case which one will can be followed? Always i will have a triangulated consensus ask value that is predisposed for showing BUY positions and a bid value that will show more often SELL signals.
Or the negate: If using Ask it would be harder to see Sell signals, and for Bid harder to see BUY signals.
Another problem - technical this time: If the EA runs on currency X and needs the values from currencies A,B,C if A,B,C changes X will not change. If X changes A,B,C values will be read.
So at time T i might have a good signal for Buy or Sell but won't be found/shown by the EA because X was not updated by the metatrader even if A,B,C changed. This way you can fail to open new positions or fail to close when would be the time or close with loss a good position.
RefreshRates doesn't solve the problem because Start is executed only when a new "feed" from currency X was received.
If this cannot be solved i'm thinking of writing my own code that uses the packets metratrader switches on LAN with the broker server. This way it could be solved your other problem of checking some data from another website.
So instead of using a platform i could make my own standalone exe that communicates with broker server and also with other servers for additional informations. Anyway have to see how it performs further, will see...
Attached is a picture of problem 1 if anyone interested to see it.
- check In the top chart is the ASK price triangulated and in the bottom one the BID - check the big differences.
Tjpld: Forgot to tell you, this ea can't be backtested only forwardtested. Backtest anyway can't be done here i would need thick charts very very accurate.The trade is not essential at this point, it was placed there just for test there are bigger problems than that trade
Ok, Have a nice weekend, i wait the next week to play with it more. Congrats for the thread/idea kurka, it set me back to work.
In an EA like this you have to use a continuous loop in order to have the correct pricing at any given time. I loop my Start function like this.. Just make sure that all your functions dont close the loop or return the start().
PHP Code
int start() { while(IsConnected()){ TickStart = GetTickCount(); //Count milliseconds for each loop for code optimization CheckBid = Bid; // set Bid to correct value for calculations for upcoming cycle MarketHours(); //Check to see what the market conditions are if(OrdersTotal() > 0){ManageOrders();} // Look for open orders and manage them Triggers();//Check to see if there are any new order opportunities ChartComment(); Sleep(10); } return(0); }
Example if you looking for signals on EURUSD according to the AUD. a buy signal would use ASK("EURUSD") < BID("AUDUSD") * ASK("EURAUD") But a sell signal would use BID("EURUSD") > ASK("AUDUSD") * BID("EURAUD"). Hope that helps you out... Finanaly got someone here to discuss and bounce ideas off.
Keep it simple stoopid....
Kurka,
Thanks for the idea kurka i was thinking more to what SwingMan said of using an EA on many charts which will save any modification of price somewhere and the one that runs on base curency to use that info. (global variables SwingMan said)
Anyway your idea seems to be easier to implement but i think it will crush metatrader if i don't close somehow the loops when another thick comes. I think i will use the volume for this: i will save static volume when entering the loop and when it changed it means a new thick was received and close the loop, will test some more to see if it works. What i want to do is to keep only 1 loop at a time. Are you using GetThickCount for that? Basicaly something that changes when new thick comes has to be used, GetThickCount also, should be ok i think.
Another thing: taking positions only in the direction of your fundamentals i think it's not very good because market can have desequilibrium in both sides of the consensus, at different moments of time, even if the main trend is in one direction. Taking positions only in one direction because of fundamentals, might loose a lot of good positions. After all we are targeting 1pip in 80% of trades, or something like that.
Anyway the TP seems to be clear, 1 pip for all trades and more for those where market sais it is still desequilibrated and will go more. For SL i think the opposite could do it for now but i'm not very sure about. Also taking fundamentals into consideration(indicators/indexes or other stuff) won't change the pure idea of trading the market based on what "Market Thinks"? A desequilibrium is something that happens all the time and it's regulated back because of logical mathematical reasons but fundamentals are speculations.
About MM for my own case i'm thinking of something very hardcore: I'll test it for a week or so and see which was the biggest drawdown i had for a given pair, i add a fiew more pips for margin and use the maximum lot availble taking those things into consideration and the current equity. This way the profit will increase exponentialy. It might break the account fast but it might also grow it exponentialy fast. I will do it anyway in demo. Basicaly i will agree to see an X% drawdown, if the drawdown i see on the testing week is smaller, i will increase the lot size until that drawdown will touch my X% of the account size.
I think the safest thing would be to see which is the biggest drawdown in pips for a period of time,...the bigger period the better and use that info and the current equity to obtain the lot size. The idea is to know that the drwadown won't go more than X% and knowing that i will use my maximum lot size for that to happen. Please let me know what you think.
Also, instead of using the whole equity for this drawdown touch thing i can divide the equity in a fiew parts and play with one part at a time as it would be the whole equity. If it doubled i take 25% and give to other parts and so on. Can be done many things if you know that the ea it's going on positive side.
Giraia_br
I can make an indicator but it's not profitable, this strategy aims most of the time for 1 pip so trading manualy using an indicator would be a had ache. Anyway you can use the colors the test ea draws to trade manualy if you realy want it.
Thanks, Victor
Thanks for the idea kurka i was thinking more to what SwingMan said of using an EA on many charts which will save any modification of price somewhere and the one that runs on base curency to use that info. (global variables SwingMan said)
Anyway your idea seems to be easier to implement but i think it will crush metatrader if i don't close somehow the loops when another thick comes. I think i will use the volume for this: i will save static volume when entering the loop and when it changed it means a new thick was received and close the loop, will test some more to see if it works. What i want to do is to keep only 1 loop at a time. Are you using GetThickCount for that? Basicaly something that changes when new thick comes has to be used, GetThickCount also, should be ok i think.
Another thing: taking positions only in the direction of your fundamentals i think it's not very good because market can have desequilibrium in both sides of the consensus, at different moments of time, even if the main trend is in one direction. Taking positions only in one direction because of fundamentals, might loose a lot of good positions. After all we are targeting 1pip in 80% of trades, or something like that.
Anyway the TP seems to be clear, 1 pip for all trades and more for those where market sais it is still desequilibrated and will go more. For SL i think the opposite could do it for now but i'm not very sure about. Also taking fundamentals into consideration(indicators/indexes or other stuff) won't change the pure idea of trading the market based on what "Market Thinks"? A desequilibrium is something that happens all the time and it's regulated back because of logical mathematical reasons but fundamentals are speculations.
About MM for my own case i'm thinking of something very hardcore: I'll test it for a week or so and see which was the biggest drawdown i had for a given pair, i add a fiew more pips for margin and use the maximum lot availble taking those things into consideration and the current equity. This way the profit will increase exponentialy. It might break the account fast but it might also grow it exponentialy fast. I will do it anyway in demo
. Basicaly i will agree to see an X% drawdown, if the drawdown i see on the testing week is smaller, i will increase the lot size until that drawdown will touch my X% of the account size.I think the safest thing would be to see which is the biggest drawdown in pips for a period of time,...the bigger period the better and use that info and the current equity to obtain the lot size. The idea is to know that the drwadown won't go more than X% and knowing that i will use my maximum lot size for that to happen. Please let me know what you think.
Also, instead of using the whole equity for this drawdown touch thing i can divide the equity in a fiew parts and play with one part at a time as it would be the whole equity. If it doubled i take 25% and give to other parts and so on. Can be done many things if you know that the ea it's going on positive side
.Giraia_br
I can make an indicator but it's not profitable, this strategy aims most of the time for 1 pip so trading manualy using an indicator would be a had ache. Anyway you can use the colors the test ea draws to trade manualy if you realy want it.
Thanks, Victor