Welcome to Forex-TSD!, one of the largest Forex forums worldwide, where you will be able to find the most complete and reliable Forex information imaginable.
From the list below, select the forum that you want to visit and register to post, as many times you want. It’s absolutely free. Click here for registering on Forex-TSD.
Exclusive Forum
The Exclusive Forum is the only paid section. Once you subscribe, you will get free access to real cutting-edge Trading Systems (automated and not), Indicators, Signals, Articles, etc., that will help and guide you, in ways that you could only imagine, with your Forex trading.
Elite Section
Get access to private discussions, specialized support, indicators and trading systems reported every week.
Advanced Elite Section
For professional traders, trading system developers and any other member who may need to use and/or convert, the most cutting-edge exclusive indicators and trading systems for MT4 and MT5.
I do not know how to cut and paste a trade from meta trader and paste on this or any forum.
But, I had a long trade signal in the 4h time frame. I had a green signal up from the AC and AO plus the Stoc. both were up on the other side of the AC zero line.
After I placed my trade it retraced a little but all my long signals were all gone. Both AC and AO had turned red pointing down and the Stoc. was also pointing down.
Because this happened so quickly is this a repaint?
Or what is this?
If you're using the standard AO/AC that comes with MT4 you can absolutely sure there is no repaint.
At 0000 GMT, USDCAD opens a new candle and a valid short signal is received from all three indicators. Five trades were opened as expected. The first trade hit TP 50 pips at 1128 GMT and all remaining four trades have their SL move to BE.
Unfortunately, the market reverses and the four trades hit SLs at 2135 GMT.
However, because of all this happens in a period of one day, the EA opens another five trades at 2135 GMT. The signal is late and therefore it should have stayed idle until the next candle at 0000 GMT.
I will fix this bug, but give me until the end of the week as I am currently involved with another project.
Azmel,
I just backtestedd the GBP/USD and I believe there are still some issues.
1) I found a '5th' trade that should have been closed with a signal from stoch/AC but was not. - see screenshot
2) I found a valid entry signal that was not taken - I believe due to a position being open when the signal was generated (that ol' chestnut). - see screenshot
3) I also believe that a 'signal is not a signal' until it is confirmed by the period closing - this means that all our trades should be initiated at the open of the next day at 0.00GMT. It appears to me that positions are being taken as soon as the AC crosses and Stock & 200sma criteria agree. This does affect the outcome - pos or neg I don't know, but it does affct the price at which we enter and may even effect whether we enter or not. This is my understanding of how to read signals generated from most indicators. It would be worth confirming with OzFx on whether or not he waits for the period to close to 'confirm' the signal is valid.
Azmel,
I just backtestedd the GBP/USD and I believe there are still some issues.
1) I found a '5th' trade that should have been closed with a signal from stoch/AC but was not. - see screenshot
2) I found a valid entry signal that was not taken - I believe due to a position being open when the signal was generated (that ol' chestnut). - see screenshot
3) I also believe that a 'signal is not a signal' until it is confirmed by the period closing - this means that all our trades should be initiated at the open of the next day at 0.00GMT. It appears to me that positions are being taken as soon as the AC crosses and Stock & 200sma criteria agree. This does affect the outcome - pos or neg I don't know, but it does affct the price at which we enter and may even effect whether we enter or not. This is my understanding of how to read signals generated from most indicators. It would be worth confirming with OzFx on whether or not he waits for the period to close to 'confirm' the signal is valid.
Cheers Jase.
Thanks Jase for the updated observation.
I checked the codes again and the equations for entering and closing trades are pretty solid. Does (1) and (2) occurs everytime? Or does it occur occassionally, say 1 out of 10, or 1 out of 100, etc? If it is the latter, then my explanation is that the backtester simply missed it. In my forward tests, everything looks fine so far, despite everything being slow because of the real-time trading, of course. We'll see in due time if the EA missed a signal in forward-test later on (we have to wait a while, though...).
Regarding (3), yes, that is how the current EA looks at it. If at 0000 GMT, there is a no-go, but a few hours later there is a valid signal, the EA makes the trade. This is the difference between manual and automatic trading. In manual trading we can't monitor the market every minute, but the EA can. I'm not sure what is the implication of doing this obviously.
I can adjust the code so that the EA will only "look once" at 0000 GMT and that's it. If there is a no-go signal at 0000 GMT, then the EA will sleep regardless what happens later in the day and wakes up again at the next 0000 GMT. This is emulate manual trading 100%
The examples I gave were the only errors I noticed for the whole backtest which spans nearly 3 years. I did re-run the backtest and obtained exactly the same result with same errors - it's quite puzzling.
The reality is that if the system rules are followed to the letter, then it should be impossible to have two positions generated from different signals open at the same time - as in order to get the new signal the closure of the previous position is necessary.
As far as only reading signals once per period - the only way to see how this would affect results would be to give it a try. It would solve the problem you encountered earlier on the USD/CAD would it not?
How difficult is it to add in some user variables so that the EA may be optimised? Perhaps variable SL, TP, Trailing Stop, Option to trade 1,2,3,4 or 5 lots with different profit levels. I'll leave it up to you, I realise it's all time and appreciate how far you have taken it so far.
Cheers, J.
If you manually backtest this system over the course of a few years, it does not turn a profit. The trending periods don't make up for the drawdowns you will encounter during choppy periods.
Anyone trading this strategy should keep that in mind.
Yes I agree with you that by letting the EA trade only at 0000 GMT will eliminate the problem I highlighted in post #92. I will make the ammendments.
aegis,
Yes, based on Jase's backtest results, it does appear so. This EA will need some optimisation on the parameters. I will make those parameters as variables for Jase to optimise and (hopefully) find the correct settings to turn this into a profitable system.
OzFx
Linear Mode
