Multi EA - page 9

[dmbsys]

Here's a copy of my screen (Demo Account)

Michel

[jlpi]

As Shinigami pointed out it is important for backtests to use a shift of 1 instead of 0 because the close is a known point but the variations inside a candle (at least M1) are just modelized by the testing tool so it may be wrong.

Of course on the latest version (with H4) if you put 1 instead of 0 the Equity will drop constantly. You can argue with reason that the shift of 1 on a H4 timeframe is too much.

So you can try on a lower time frame, If you take shift = 1 on M1 and for RSX you use 720 instead of 3 (3 x 240 min). And you will also get a steady decreasing equity.

Maybe my use of RSX is wrong but I think that as long as we cannot see profits with a version with a shift of 1 we can consider that this EA is not profitable.

[dmbsys]

Ok jlpi,

Try this ; I've modified the EA to put a 'ShiftValue' and 'ReverseCondition'.

With a shift = 1, I reverse condition... also on timeframe H4.

Look backtest.

Michel

[Tarpan]

Merci beaucoup Michel.

I will test your version.

[[Deleted]]

[Tarpan]

jlpi:
As Shinigami pointed out it is important for backtests to use a shift of 1 instead of 0 because the close is a known point but the variations inside a candle (at least M1) are just modelized by the testing tool so it may be wrong.

Of course on the latest version (with H4) if you put 1 instead of 0 the Equity will drop constantly. You can argue with reason that the shift of 1 on a H4 timeframe is too much.

So you can try on a lower time frame, If you take shift = 1 on M1 and for RSX you use 720 instead of 3 (3 x 240 min). And you will also get a steady decreasing equity.

Maybe my use of RSX is wrong but I think that as long as we cannot see profits with a version with a shift of 1 we can consider that this EA is not profitable.

Try to backtest in 30 min. using shift 1 and reverse condition

[jlpi]

dmbsys:
Ok jlpi,

Try this ; I've modified the EA to put a 'ShiftValue' and 'ReverseCondition'.

With a shift = 1, I reverse condition... also on timeframe H4.

Look backtest.

Michel

Thank you for this new version that looks more interesting.

It is interesting to note that changing the shift from 0 to 1 you had to reverse the orders !

Hello FerruFx, nice to see you here as a very active member !

[Chrisstoff]

Hi dmbsys,

Thank you for your modifications. I have been testing your first 2.0.5 version (actually, the current version would not have been numbered as 2.0.6?) from this morning only on GBPUSD. There were 5 trades so far, the total result is +8 pips.

It seems to be promising

Tomorrow I will try the current version as well, probably on all the three pairs.

Regards,

Chrisstoff

dmbsys:
Ok jlpi,

Try this ; I've modified the EA to put a 'ShiftValue' and 'ReverseCondition'.

With a shift = 1, I reverse condition... also on timeframe H4.

Look backtest.

Michel

[dmbsys]

chrisstoff:
Hi dmbsys,

Thank you for your modifications. I have been testing your first 2.0.5 version (actually, the current version would not have been numbered as 2.0.6?) from this morning only on GBPUSD. There were 5 trades so far, the total result is +8 pips.

It seems to be promising

Tomorrow I will try the current version as well, probably on all the three pairs.

Regards,

Chrisstoff

Sorry for the numbered...

But try too on USD/JPY : work too !!!

I test it on 4 pairs : EUR/USD, USD/CHF, GBP/USD & USD/JPY

Michel

[dmbsys]

Hi,

I've modified the version (for the last time, I hope).

Here's :

1. I use the time (UseHourTrade=true)

2. Modification of condition to close order :

int CloseTrade()

{

for(int i=0;i<OrdersTotal();i++)

{

if(OrderSelect(i,SELECT_BY_POS,MODE_TRADES)==false) break;

if(OrderSymbol()==Symbol() && OrderMagicNumber()==MagicNumber)

{

if(OrderStopLoss()==0)

{

if(OrderType()==OP_BUY) OrderClose(OrderTicket(),OrderLots(),Bid,3,Yellow);

if(OrderType()==OP_SELL) OrderClose(OrderTicket(),OrderLots(),Ask,3,Yellow);

}

}

}

}

Michel