Miscellaneous Dumb Questions – All EAs

[crazzzzzy]

What are magic numbers and what does changing the magic number do :

[garyfritz]

It's an identifier you can attach to open orders. Then you can write EAs / scripts to only pay attention to particular magic numbers. Not all EAs support this.

[ahniel]

Hi there

Could someone please enlightened me, in a live trading with Nanningbob EA, what happens if I turn of the computer/disconnect with live trades.

Would my TP/SL be hit or must I have the computers turned on?

Thank you

[NeoTrader]

hi ahniel

Could someone please enlightened me, in a live trading with Nanningbob EA, what happens if I turn of the computer/disconnect with live trades.

Would my TP/SL be hit or must I have the computers turned on?

You shouldn't even think about it to going live with any EA if you are that fresh.

But to answer the question...all standard SL and TP settings are set on the broker server.
If you use some kind of stealth method ( I don't know if your EA does that) then it would possibly be fatal since the real SL and TP are not set on your server.

To gain more insight you should ask in the Thread of the EA.

happy trading,

NeoTrader

-

[ahniel]

hi ahniel

Could someone please enlightened me, in a live trading with Nanningbob EA, what happens if I turn of the computer/disconnect with live trades.

Would my TP/SL be hit or must I have the computers turned on?

You shouldn't even think about it to going live with any EA if you are that fresh.

But to answer the question...all standard SL and TP settings are set on the criminal server.
If you use some kind of stealth method ( I don't know if your EA does that) then it would possibly be fatal since the real SL and TP are not set on your server.

To gain more insight you should ask in the Thread of the EA.

happy trading,

NeoTrader

-

Hi NeoTrader

You are right, it is too early for me to go into live trading.

I have always been placing demo trades on Oanda Java Platform but I'm now learning how to demo trade on Empty4 instead of just using it as a chart and indicator platform.

Normally, how many months do we test an EA before going live with it?

Thanks for your enlightenment.

ahniel

[NeoTrader]

hi ahniel,

You are right, it is too early for me to go into live trading.

Thanks...for one moment I really thought you mean it serious.

I have always been placing demo trades on Oanda Java Platform but I'm now learning how to demo trade on Empty4 instead of just using it as a chart and indicator platform.

Learn all the quirks of Empty4 and your criminal and how they behave on different market conditions...this is the way to go.

Normally, how many months do we test an EA before going live with it?

There is no single timeframe to answer this question. It depends on how stable and reliable your EA is. If there are still bugs or you enhance it with new features the test cycle starts again.

If your EA trades reliably over some time and in different market conditions and you know exactly what the EA is doing and that he is doing this correct than it could be the time to go live with a small lotsize and a small account. If this also works just stock up your capital step by step...

But you also should know exactly how the strategy behind the EA works.

Thanks for your enlightenment.

This has nothing to do with enlightenment... It is plain and simple logic


happy trading,

NeoTrader

-

[nolube]

Hey there, this is my first post here so thought I'd kick off in the newbie section. I have been writing some EA's for the last 5 - 6 months with the help of Molanis Strategy Builder. I saw Steve mention elsewhere he did not want backtested results from Empty4 (assuming thats what Empty4 is lol) because the results can be innacurate. I have read this before I think in other places but basically what I wanted to ask was does this still apply (as much) if your EA trades off H4 charts? Or do the innacuracies arise only on smaller TF EA's and hence, are negligible when you are trading bigger moves?

Cheers
nolube

[NeoTrader]

hi nolube,

Hey there, this is my first post here so thought I'd kick off in the newbie section. I have been writing some EA's for the last 5 - 6 months with the help of Molanis Strategy Builder. I saw Steve mention elsewhere he did not want backtested results from Empty4 (assuming thats what Empty4 is lol) because the results can be innacurate. I have read this before I think in other places but basically what I wanted to ask was does this still apply (as much) if your EA trades off H4 charts? Or do the innacuracies arise only on smaller TF EA's and hence, are negligible when you are trading bigger moves?

The Strategy Tester in Empty4 is inaccurate from his concept. Since Empty4 does not save/store tick values the strategy tester is not really able to do a real tick based testing...this means that it simulates/emulates the ticks...and this is the reason why it is crap over all TFs...

You can still do your tests....to try to get a rough estimate...but don't see them as reliable results.

happy trading,

NeoTrader

-

[garyfritz]

I believe, however, that this affects low TFs more than high TFs. If you're using M1 data and simulating tick behavior inside candles, then testing on M1 candles will be pretty inaccurate. You have no idea what happened inside the M1 candles. But if you're running on H1 candles, there are 60 M1 "ticks" inside each H1 candle, so you can see the way price moved inside the H1 candle. (Which can be important when deciding which stop gets hit first, etc.) It becomes much closer to reality.

That's my understanding anyway. Make sense to you, Neo?

[NeoTrader]

hi gary,

I believe, however, that this affects low TFs more than high TFs. If you're using M1 data and simulating tick behavior inside candles, then testing on M1 candles will be pretty inaccurate. You have no idea what happened inside the M1 candles. But if you're running on H1 candles, there are 60 M1 "ticks" inside each H1 candle, so you can see the way price moved inside the H1 candle. (Which can be important when deciding which stop gets hit first, etc.) It becomes much closer to reality.

That's my understanding anyway. Make sense to you, Neo?

You are absolutely right in your statement and it makes absolute sense Gary...
But the problem still remains that in volatile pairs in one minute can happen pretty much movement...especially when news related...and there is no way in Empty4 to catch/record this...and to use it for testing....

Therefore any attempt to do a reliable and realistic testing that is near the realtime market movement is futile in the Empty4 environment and is meant to fail...

If you have a strategy that uses candle close/open for the entry then it could be somewhat accurate but anything else is nothing more than a rough estimate....

I'm looking into this .net api thingy that bridges to Empty4....if there is any way that I can use this to feed realtime market data (from ducascopy) into the indicators inside Empty4 and can get their calculations out to the .net application...this would be a potential base to build a reliable testing platform...
There is already a somewhat similar Open Source Project called Forex Strategy Builder(FSB)/Forex Strategy Trader(FST) ... but they don't really have a asynchronous bridge. You must develop all indicators on their Platform...

happy trading,

NeoTrader

-