The Great HGB Foward Test

Leslie · Post by **Leslie** » Tue Jan 13, 2015 8:05 pm

I'm a newbie and in the US with stupid FIFO and stop/limit netting but would love to participate- no VPS but will run 24/5.

Home Computer (1)

Leslie

fxpoison · Post by **fxpoison** » Tue Jan 13, 2015 8:19 pm

Yep me too, I have a 24x5 setup.
Let's start "musing" and attracting inputs about the settings to test.

FXe3 · Post by **FXe3** » Tue Jan 13, 2015 9:15 pm

FYI, ATC Brokers takes care of all the FIFO and hedging on the backend. No need to worry about FIFO and hedging violations

"In order to abide by the CFTC FIFO rule, the counter party offsets the positions on a First In First Out basis. On the Empty4 4 platform you have the ability to close the positions as you see fit. For example, if you buy a contract and buy another contract at a later time and you decide to close your most recent buy position, you can simply close this position on the Empty4 interface, however, the order that is sent to the counter party will be offset with the FIFO rule in place. Your net position remains the same on the Empty4 and the back office. For the Trading Station II platform, FIFO is applied only to US accounts."

Leslie » Tue Jan 13, 2015 8:05 pm wrote:I'm a newbie and in the US with stupid FIFO and stop/limit netting but would love to participate- no VPS but will run 24/5.

Leslie

Post by **trendinator** » Wed Jan 14, 2015 1:15 am

Thanks for your interest guys!

If you have already posted, can you please edit your existing post with the number of instances of Empty4 you are able to run and what computer it will run on. eg. Amazon VPS, Another VPS (please state), or home/office computer.

If you are going to use the Amazon free VPS, I have tested it and it will run ONE Empty4 with HGB & 28 pairs. Unfortunately the Amazon VPS cannot handle more than one Empty4 at a time.

My example:

Amazon VPS (1)

I will continue updating the first posts with more info. Currently waiting on version 3 of HGB which will hopefully have all the bugs ironed out.

Post by **boinvest** » Wed Jan 14, 2015 2:49 am

trendinator » Wed Jan 14, 2015 1:15 am wrote: I will continue updating the first posts with more info. Currently waiting on version 3 of HGB which will hopefully have all the bugs ironed out.

well I have to disappoint you, with v3 I am taking Steves she'll to the next level. Thougt it was quicker than starting my own shell thingy from scratch. Well still think the approach I choose is quicker.
But it is a major rewrite. One of the biggest changes is the order accounting function.
I am about to incorporate the library version of HGB into it. According to Milanese it will improve performance a lot. The new accounting function should also reduce system resources.
Be prepared for a beta version which wont be stable.

I was about to make a plan for demoing this bot. But so busy with coding that I didn't get to work out the plan or start a lot of demos. Have some ideas for modded versions to test some concepts. Well didn't get to that either so won't bother you with it.

Great stuff a collective forward test. As there are so many options we need a good plan.
Well first thing to make it easier to tear this thingy apart.
Have only one type of trading on the account for easy comparison.
So only trend with all others disabled. Then compare some accounts with different SL and TP 's.
Use post 1 to show all fxbook widgets..
For a start focus on 1 timeframe, the H4 because thats what it's designed for.
Then use H4 as medium Tf and different lower, to see what that brings.
Further on use H4 trading and some different higher time frame.
After that, put them all together? All Tf filters used.

Another great idea, don't use personal my fxbook, but make one account for each trading type. The portfolio function is very useful for determining the best set

Well have more ideas, those are for later.
I also gathered every pc which I could lay my hands on, with the plan to run a lot of demo's.
I wanted to use a lightweight Linux distro like LXLE on USB stick, which allowed me hopefully to run more instances then Windows. Plus on some of the PCs are fresh Windows install which I want to leave that way.
Maybe we can make a collective traders USB stick OS

OK back to testing, at this moment I think performance of range trades could benefit when close on opposite tma is set to false. Just spotted this one in the code. Could some one test this and confirm?
Would also be great to pay some attention to the atr and jumping candle and trailling stops.

Enough to play with

FXe3 · Post by **FXe3** » Wed Jan 14, 2015 3:04 am

Count me in guys. I could run 4 or 5 instances, maybe more
Home PC

Post by **trendinator** » Wed Jan 14, 2015 3:35 am

boinvest wrote:
Have only one type of trading on the account for easy comparison. So only trend with all others disabled.
Then compare some accounts with different SL and TP 's.
Use post 1 to show all fxbook widgets..
For a start focus on 1 timeframe, the H4 because thats what it's designed for.
Then use H4 as medium Tf and different lower, to see what that brings.
Further on use H4 trading and some different higher time frame.
After that, put them all together? All Tf filters used.

Another great idea, don't use personal my fxbook, but make one account for each trading type. The portfolio function is very useful for determining the best set

Thanks for your input boinvest! i agree with all of your points quoted above.

I think it may be interesting to explore smaller timeframes as well, eg. H1 or even M30. but they are not a priority.

I think the ATR SL / TP may have some potential and could be explored. My understanding is that it will be able to adjust to the market conditions better?

Let's please keep all bug reports and HGB related matters in the http://www.stevehopwoodforex.com/phpBB3 ... =87&t=3966 thread.

The USB stick OS is an interesting idea, please keep us posted. Linux ey?

Looking forward to V3!

Post by **trendinator** » Wed Jan 14, 2015 3:53 am

Which pairs to test... EURCHF is pretty much a flatline, so no point trading that right? so 27 pairs?

XAUUSD could show some potential but i think we should test it separately because it will skew the pip results. (10 gold pips = 1 regular pip in terms of profit per 0.1 lot)

Post by **boinvest** » Wed Jan 14, 2015 12:33 pm

trendinator » Wed Jan 14, 2015 3:53 am wrote:Which pairs to test... EURCHF is pretty much a flatline, so no point trading that right? so 27 pairs?

XAUUSD could show some potential but i think we should test it separately because it will skew the pip results. (10 gold pips = 1 regular pip in terms of profit per 0.1 lot)

Well, thing is, EURCHF could also be interesting to trade.
As this bot, can trade in Trend and ranging markets.
And the range trades would be able to grasp some pips from these markets.

This has to do with how the EA deals with the Range trades.
And how the indicator is painting them.
Not really anything to be discussed in detail in this topic.
Just wanted to share that I am discussing this with other HGI trades, how to implement these signals.
This is one of the thingys to be improved in the HGB.

I agree, not to incorporate the Commoditys in the test..
HGB COULD work on them, and maybe preform very wel.
But the HGI system is not based on it. Bob analysis of the majors resulted in the HGI.
So better stick to majors for the demos.

Post by **boinvest** » Wed Jan 14, 2015 12:36 pm

FXe3 » Wed Jan 14, 2015 3:04 am wrote:Count me in guys. I could run 4 or 5 instances, maybe more

Fx3e You posted something about EA optimization in the HGB thread.
I cannot find it anymore. Too long did'nt read it all.

But it looked very interesting,
Over here seems to be a more appropiate place, Could you share it??